Training-Jun.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data = '/mnt/dataset/train-valid_set'\n",
    "inference_data = '/mnt/dataset/test_set'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import matplotlib.pyplot as plt\n",
    "from PIL import Image\n",
    "import numpy as np\n",
    "import math\n",
    "\n",
    "file_list = os.listdir(train_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Suction Irrigator (SIRL-UCUK)',\n",
       " 'R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       " 'Needle Holder (AESCULAP) (NDHL-UCAE)',\n",
       " 'Graspers (straight) (OLYMPUS) (GRSL-UCOL)',\n",
       " 'Metal Clip Applier (Medtronic) (CLAL-M3MT)',\n",
       " 'R-Maryland (da Vinci) (BFCR-MADV)',\n",
       " 'R-LND (da Vinci) (NDDR-LADV)',\n",
       " 'Prograsp (da Vinci) (GRSR-UCDV)',\n",
       " 'Polymer Clip Applier (CLAL-UCUK)']"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "file_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "class_list = ['Prograsp (da Vinci) (GRSR-UCDV)',\n",
    "              'R-LND (da Vinci) (NDDR-LADV)',\n",
    "              'R-Maryland (da Vinci) (BFCR-MADV)',\n",
    "              'R-Scissors (da Vinci) (MSCR-UCDV)',\n",
    "              'Suction Irrigator (SIRL-UCUK)',\n",
    "              'Graspers (straight) (OLYMPUS) (GRSL-UCOL)',\n",
    "              'Needle Holder (AESCULAP) (NDHL-UCAE)',\n",
    "              'Metal Clip Applier (Medtronic) (CLAL-M3MT)',\n",
    "              'Polymer Clip Applier (CLAL-UCUK)'\n",
    "             ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "\n",
    "def getjson(json_path):\n",
    "    with open(json_path) as json_file:\n",
    "        json_data = json.load(json_file)\n",
    "    return json_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "12408"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 근데 이미지 하나당 수술도구 하나씩 나오니까\n",
    "# 처음부터 데이터를 전부 읽어와서 리스트 하나로 만들어봤어\n",
    "\n",
    "# all_data 는 모든 json file 리스트\n",
    "# data_dic 는 key 값이 수술도구, value가 json file인 딕셔너리\n",
    "\n",
    "all_data = []\n",
    "data_dic = {}\n",
    "\n",
    "for tool in class_list:\n",
    "    files = os.listdir(train_data+'/'+tool)\n",
    "    json_files = [file for file in files if file.endswith(\".json\")]\n",
    "    \n",
    "    all_data.extend(json_files)\n",
    "    data_dic[tool] = json_files\n",
    "\n",
    "len(all_data) # 모든 json file은 12408개"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1622\n",
      "663\n",
      "1962\n",
      "3016\n",
      "1480\n",
      "805\n",
      "978\n",
      "789\n",
      "1093\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "12408"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 근데 이미지 하나당 수술도구 하나씩 나오니까\n",
    "# 처음부터 데이터를 전부 읽어와서 리스트 하나로 만들어봤어\n",
    "\n",
    "# all_data 는 모든 json file 리스트\n",
    "# data_dic 는 key 값이 수술도구, value가 json file인 딕셔너리\n",
    "\n",
    "all_data = [[None]] * 9\n",
    "\n",
    "for idx, tool in enumerate(class_list):\n",
    "    files = os.listdir(train_data+'/'+tool)\n",
    "    json_files = [file for file in files if file.endswith(\".json\")]\n",
    "    \n",
    "    all_data[idx] = json_files\n",
    "    \n",
    "count = 0\n",
    "for i in range(9):\n",
    "    count += len(all_data[i])\n",
    "    print(len(all_data[i]))\n",
    "# 모든 json file은 12408개\n",
    "count"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_data_in = []\n",
    "data_dic = {}\n",
    "\n",
    "for tool in file_list:\n",
    "    files = os.listdir(train_data+'/'+tool)\n",
    "    json_files = [file for file in files if file.endswith(\".json\")]\n",
    "    \n",
    "    all_data_in.extend(json_files)\n",
    "    data_dic[tool] = json_files\n",
    "    \n",
    "all_data_in = list(set(all_data_in))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'SNUH_DC07_JCW0_RLPN_0002_0078956.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00196878.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00030546.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0002_00037459.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0009_0003143.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_0153210.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00052282.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0037_00223539.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00027550.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0002_00006753.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00038969.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00060403.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00061225.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0099490.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0056_00278038.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0057_00122350.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00105793.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00063787.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0007_00026115.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_0167856.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0020_0000508.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00066201.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0006_00042209.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0013_00018586.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_0086463.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0030211.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0008_00082359.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0008_00076142.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0014_0014205.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00084946.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00099607.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0013_00138805.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0004_00015807.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0071134.json': ['Suction Irrigator (SIRL-UCUK)',\n",
       "  'R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00024473.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0022_00089565.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00084383.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0017_00021942.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0050525.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00250201.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00177052.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0046_00284637.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0017_00019600.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0002_00034408.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0086925.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00037788.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00118080.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00047076.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00040504.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00037816.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00077084.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00074583.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0038_00192796.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0007_00026517.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0082_0160823.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0073078.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC16_RSB0_RHC0_0010_00136424.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0011_00069680.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00015627.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00016719.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC16_PJW0_RHC0_0007_00154955.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0047307.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0008_0039348.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00018934.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00198300.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_0087547.json': ['R-Maryland (da Vinci) (BFCR-MADV)',\n",
       "  'Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00021845.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0042_00241711.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_0128250.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00074141.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00035453.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0007_00026041.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0014_0026298.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0008_0088699.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0050_00292440.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC16_KMJ0_RHC0_0001_00050902.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0005_00050928.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC16_PJW0_LAR0_0008_0019869.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00131858.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00106085.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00193508.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0014_0025779.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0051658.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_0037473.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC16_KMJ0_RHC0_0001_00067010.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0035_00327887.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0022_00159012.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00097180.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00090878.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0009_00038573.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0044587.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0031_0366466.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00021860.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00068689.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0007_00026077.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0047315.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0007_00030644.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_0006844.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0013_00087331.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0046373.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0053_00096505.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00036513.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_0113336.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0048_00115533.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0009_0002461.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00074248.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00070209.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00097750.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00051460.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0020_0013537.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00032482.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0037_00206013.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00070202.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00039849.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0057813.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00078927.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'GGHB_DC68_LJH0_BABA_0003_0003212.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0008_0084199.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0038431.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0012_00021000.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC68_LKE0_BABA_0002_0306287.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0127418.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00016706.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC16_RSB0_RHC0_0003_00041350.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00071025.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00059513.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00089839.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC16_RSB0_RHC0_0010_00136798.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0047171.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0030999.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0022_00089551.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_0009654.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00075223.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00052287.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0040_00024096.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0033261.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00186158.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00138929.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0020546.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0046_00273852.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00038547.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00038431.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00062941.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC16_KMJ0_RHC0_0001_00067306.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_0007762.json': ['R-Maryland (da Vinci) (BFCR-MADV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00187882.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0097594.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00043760.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00010334.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0011_00072434.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0007_00030184.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00098526.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0004_00051890.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0100582.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00074486.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_0013793.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00032477.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0008_0021265.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00066043.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00051394.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00036524.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00012115.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00009136.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0012901.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0074880.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00193162.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC16_PJW0_RHC0_0007_00153075.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00059649.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00044750.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0014_0006073.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00265820.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00062509.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00119161.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0044682.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)',\n",
       "  'Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0023_00029918.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0013_00138898.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00062009.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00062245.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0007_00056364.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0011_00075590.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00016381.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0006_00027646.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00043288.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00021917.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00047090.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_0016960.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00213282.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0014_0015115.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00021951.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0057767.json': ['Suction Irrigator (SIRL-UCUK)',\n",
       "  'R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00018120.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00025507.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0023_00029633.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0002_00059149.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0009_0078985.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0008_0043287.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00070624.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00050576.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00074292.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0008_0023889.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0020_0001480.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00007178.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0020_0006462.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00050794.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00069783.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0054_00187130.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0008_00049306.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_0059125.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00077806.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00009129.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0097961.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00019043.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0008_00082399.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00050374.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0004_00041464.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00074322.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0062539.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00071034.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00068298.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00100014.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00151440.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0008_0088879.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00059528.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0008_00039841.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC16_RSB0_RHC0_0003_00041040.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'GGHB_DC68_LJH0_BABA_0003_0017961.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC16_KMJ0_RHC0_0001_00067034.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00045636.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0062161.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00102558.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0012074.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Graspers (straight) (OLYMPUS) (GRSL-UCOL)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0030811.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00071861.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0002_00012154.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0048_00190081.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0046444.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0032_0243714.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00169814.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00053461.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00069361.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00012007.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0020_0031486.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00049953.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00060037.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0009738.json': ['R-Maryland (da Vinci) (BFCR-MADV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0040_00275411.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_0016217.json': ['Suction Irrigator (SIRL-UCUK)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0036_00156638.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC16_PJW0_RHC0_0001_00120093.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00084322.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0024029.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00035663.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00021774.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0006_00042458.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0004_00030240.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00052265.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00021760.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0014_0005721.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0031_0130792.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00078836.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC16_PJW0_LAR0_0010_0020268.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00057005.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0031_0178033.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0023_00029670.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0035131.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00098569.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00074422.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00183774.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0044344.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0064309.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0036461.json': ['Suction Irrigator (SIRL-UCUK)',\n",
       "  'R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0037_00244509.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0048576.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0012921.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0023_00039697.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00078841.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0036_00243240.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00018195.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00201133.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00070983.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0009_0078861.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00039181.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00082480.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00062509.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00064066.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0020_0089851.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0035_00429810.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0008_00011187.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0031_0360864.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0035671.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00010282.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00169943.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00091790.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0124458.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC16_KMJ0_RHC0_0001_00051122.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00037897.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00080520.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00052285.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0073002.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00074496.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0046280.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0054695.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00077042.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0006_00027081.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00045617.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0090530.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC16_PJW0_LAR0_0008_0019893.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0066067.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0002_00012463.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00222046.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0002_00066245.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00074305.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00186154.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00006814.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00265843.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0004_00052112.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0061311.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0102537.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00074253.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0002_00007015.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'GGHB_DC68_LJH0_BABA_0002_0015121.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00063166.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0008_0080599.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00018682.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0008_00082287.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00083922.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00063943.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00066081.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00060020.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00025008.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00099572.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00063834.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0046931.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00060942.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00059530.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00071865.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0012_00008074.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00155209.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0014_0014287.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00021890.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00067600.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00061473.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00193160.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00068577.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0082_0142008.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0032191.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00183640.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00098578.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00058534.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC16_PJW0_LAR0_0010_0019608.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0009_0002835.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0133061.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00056590.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00024692.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00001787.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0011140.json': ['Suction Irrigator (SIRL-UCUK)',\n",
       "  'R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00063895.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0046360.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0051508.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0031_0002861.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00070225.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00186151.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00064105.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0054036.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_0001338.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0089845.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00068534.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00050314.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0014_0042504.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC68_LKE0_BABA_0001_0132257.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC16_PJW0_LAR0_0004_0028152.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00017762.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00010778.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0013_00082271.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0006_00021818.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0043_00212044.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0019037.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0013_00009239.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC68_LKE0_BABA_0001_0145840.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0059957.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0005_00050962.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00081778.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0050836.json': ['R-Maryland (da Vinci) (BFCR-MADV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)',\n",
       "  'Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00075225.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00165958.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_0133380.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0008_00058728.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_0011000.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00028329.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0090500.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0036315.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00017085.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00133119.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0054_00225303.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0040265.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)',\n",
       "  'R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00064784.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_0037472.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00098923.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0009_0010499.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00063852.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00042829.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0082456.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)',\n",
       "  'R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00010374.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0038286.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0014_0032821.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0009_0010561.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00016690.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00064260.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0050201.json': ['Suction Irrigator (SIRL-UCUK)',\n",
       "  'R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0039_00295847.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0008_0087799.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0038_00202206.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00132946.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00037798.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00055060.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0050005.json': ['R-Maryland (da Vinci) (BFCR-MADV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00050190.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00132966.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_0197990.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC16_PJW0_LAR0_0010_0016785.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00009092.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0061071.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0010852.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_0014289.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0012_00022302.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00069558.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00049107.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0079216.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0047091.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00050717.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'GGHB_DC68_LJH0_BABA_0002_0012830.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00072871.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0036_00130516.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00060115.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0007_00026147.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0036_00135192.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00019277.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00059501.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0138038.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0081836.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)',\n",
       "  'Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0087358.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0002_00022272.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00039217.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0013781.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0138508.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0008_00029005.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC16_RSB0_RHC0_0005_00047653.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0023_00030952.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0002_00006999.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00018164.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00010666.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0053_00260761.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC16_RSB0_RHC0_0003_00040645.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0076726.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00098582.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0040_00336750.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00083233.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0100833.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00071875.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0020_0037000.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0000497.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0006_00013690.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00010332.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_0026048.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0008_0084079.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00060086.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00042917.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00070677.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00030074.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00115137.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0040_00298694.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0019_00034768.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00037885.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0069271.json': ['Suction Irrigator (SIRL-UCUK)',\n",
       "  'R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Graspers (straight) (OLYMPUS) (GRSL-UCOL)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00064100.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0011_00024558.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00020715.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0055882.json': ['Suction Irrigator (SIRL-UCUK)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00053480.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00053049.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00067060.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0062871.json': ['R-Maryland (da Vinci) (BFCR-MADV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0109377.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0090460.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_0019374.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0006_00027143.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00186179.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00107681.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0025951.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0014_0023990.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0005_00070878.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00018829.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00151416.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00066908.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0043411.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0051888.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00023100.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0074300.json': ['Prograsp (da Vinci) (GRSR-UCDV)',\n",
       "  'Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0075040.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00071682.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00058081.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0006_00027170.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00062876.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0032_0243703.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0019_00098369.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0002_00034403.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00018974.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0006_00019747.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0009_0075118.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC16_PJW0_LAR0_0010_0017150.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0009822.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00000121.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00064696.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0045632.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00010146.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0002_00043208.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0002_00034484.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0050925.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0049780.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0022_00158971.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00071036.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0005_00051231.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00096165.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0034_00241695.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00063979.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00207777.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00165360.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0049_00290749.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0036_00140265.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00193159.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0081179.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00108530.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00021771.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0020_0055733.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0037725.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00016928.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0046821.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0014_0020287.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0056907.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00059741.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00075583.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0008_0087919.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0012961.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0080427.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC16_RSB0_RHC0_0010_00136465.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00069000.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0006_00079050.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00019544.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0022_00089575.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0064549.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0008_0039228.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00193125.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0007_00030561.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00027616.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0047_00166920.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00065359.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00110203.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00083305.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0055_00408929.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00043867.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0023190.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00118560.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00062827.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_0010022.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0045_00374827.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00193770.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00210767.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00068475.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00041269.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00198365.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0055_00406371.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0037_00206193.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0005_00070851.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0044891.json': ['Suction Irrigator (SIRL-UCUK)',\n",
       "  'R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00133080.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00019304.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00086975.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00012126.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00020148.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0038671.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00020259.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00075105.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0054_00187080.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00021239.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00265862.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0040891.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00020433.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00060013.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00016996.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0005_00071066.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00217524.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0009_0033024.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00019299.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0008_0025235.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0050680.json': ['R-Maryland (da Vinci) (BFCR-MADV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_0144270.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0009_0075923.json': ['R-Maryland (da Vinci) (BFCR-MADV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00074505.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0040260.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)',\n",
       "  'R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00058505.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0054_00187680.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0079336.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0048379.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0011135.json': ['Suction Irrigator (SIRL-UCUK)',\n",
       "  'R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00068581.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0037296.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0041_00074100.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0051_00205486.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00044579.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_0017639.json': ['Suction Irrigator (SIRL-UCUK)',\n",
       "  'R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00042278.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC16_RSB0_RHC0_0010_00137776.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC68_LKE0_BABA_0001_0151062.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0002_00006286.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00183736.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00015220.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0059924.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00050253.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0014_0025880.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00035903.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00035882.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00042748.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0002_00034455.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0004_00030242.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00060041.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0075783.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00151449.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0090540.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00011990.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00092518.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0020_0040975.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00023467.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00064346.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0013_00017708.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0033_0216120.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0040_00024187.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00059150.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0042320.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0012_00008010.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0014_0023829.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00074252.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0019466.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_0017578.json': ['Suction Irrigator (SIRL-UCUK)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0037_00244959.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0023_00030984.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0079116.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0013_00017770.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC68_KSJ0_BABA_0001_0227343.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00089972.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00058525.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_0110660.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00066127.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0007_00030261.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0053_00249094.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0036_00034240.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00012003.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0032_0243404.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00086974.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0007_00057511.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00186147.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0048_00116180.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00061383.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0017_00020140.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00063056.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0043651.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00065900.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0076448.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00017938.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC68_LKE0_BABA_0002_0255637.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00069819.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'GGHB_DC68_LJH0_BABA_0003_0003461.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0009_00038869.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00118620.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0074550.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00083792.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00053623.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00069791.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0025674.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0074100.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0013_00082406.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00058940.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0049917.json': ['Suction Irrigator (SIRL-UCUK)',\n",
       "  'R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0008_0022225.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0057_00189536.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0082476.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)',\n",
       "  'R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00068736.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0049248.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00106421.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00016792.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00183644.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0008_0088219.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0013711.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00063919.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC68_KSJ0_BABA_0001_0088279.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00183642.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0132542.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00042994.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_0018047.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00058481.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00026313.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0008_0022105.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0059186.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00019918.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00068668.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0002_00054953.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0099502.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00069787.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0020_0021858.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC16_RSB0_RHC0_0003_00041062.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0014_0020958.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0018746.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0004_00051875.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00038258.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0008_0035857.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00071884.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00009131.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00070719.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0013_00082422.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0048336.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00198312.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_0086523.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0009_0018282.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00075314.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0006_00017984.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0009583.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC68_LKE0_BABA_0002_0390376.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0041_00074120.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0050435.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0056_00021252.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00086985.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00066925.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00058880.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00168521.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0004_00087540.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_0021314.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0040_00275773.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00012022.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0013721.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0047742.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0052_00023713.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0007_00031269.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00003842.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'GGHB_DC68_LJH0_BABA_0003_0002717.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00019398.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC16_RSB0_LAR0_0004_00023520.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0009_0000000.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0002_00012596.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00016709.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00192042.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0048_00116157.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC16_RSB0_RHC0_0010_00136415.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0036_00151760.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_0012557.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00028317.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0055584.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0058122.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00074246.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00151463.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0035_00423221.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00074509.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0054_00187662.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00051514.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0053_00260581.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00024439.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00051280.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00151404.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC68_LKE0_BABA_0002_0302251.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00084942.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_0018322.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0067811.json': ['Suction Irrigator (SIRL-UCUK)',\n",
       "  'R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0040_00337926.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00074520.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00026654.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00071940.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00011996.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00068419.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0013501.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0061601.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0004_00052282.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00118289.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0038_00191614.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0009_0018469.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00032032.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC16_PJW0_LAR0_0010_0016821.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00057110.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00131842.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0046393.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_0011183.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00056785.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00046704.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'GGHB_DC68_LJH0_BABA_0003_0003027.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0053518.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00077082.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0074250.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)',\n",
       "  'Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00021948.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00016726.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0059082.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0004_00013221.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0038071.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0009_00038732.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00059183.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00018461.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0022_00158975.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00068472.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0020_0059563.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00044231.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00024271.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'GGHB_DC68_LJH0_BABA_0003_0022686.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00165281.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00018481.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00094939.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00098649.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0035_00223520.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00059295.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0002_00006847.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00064137.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0008_00075821.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00166019.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0048_00168285.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_0116971.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0079146.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0051078.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0014_0022800.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00075150.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00183651.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00024054.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00068551.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00064506.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00074549.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0035_00073903.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0006_00024134.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_0024677.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00010142.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0005_00070772.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00061387.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00021160.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00032898.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00164193.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0064145.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00039689.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00078915.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0020_0073376.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00045122.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0022_00158973.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0074510.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0050883.json': ['R-Maryland (da Vinci) (BFCR-MADV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)',\n",
       "  'Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00170049.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0011_00075020.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0043165.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00023528.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0022_00158970.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0008_00030257.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00015264.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00079733.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00064756.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0086945.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0059188.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_0128220.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0025251.json': ['Suction Irrigator (SIRL-UCUK)',\n",
       "  'R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0009_0078923.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00088082.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00089256.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00023541.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0047853.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00058551.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00075083.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00068544.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0007_00026050.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0028324.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0004_00038415.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00048004.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_0140700.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0015931.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC16_RSB0_RHC0_0001_00065753.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00048030.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0048516.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00100016.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0013_00018649.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC68_LKE0_BABA_0002_0272177.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0015_00029898.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0061321.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0010_00017863.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00035243.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0022_00089244.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0036_00034235.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00016434.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0047_00164055.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0046446.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0004_00051727.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0041_00075445.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00100368.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0008_0035237.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00012407.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0035_00223200.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00003158.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00198311.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0127543.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0045_00030674.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00120124.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0036_00243150.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0002828.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0019_00028725.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0094795.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0040_00025454.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_0019436.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00092880.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00059664.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0009_0080502.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0047836.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00044592.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00129665.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC16_PJW0_LAR0_0008_0019883.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0054_00188055.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00105997.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00023695.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0022_00158941.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0049778.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00068425.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0023949.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC68_LKE0_BABA_0002_0304949.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00193150.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0022_00082670.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC16_PJW0_LAR0_0010_0017167.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0017_00020818.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00026384.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00093246.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0087661.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00164862.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC68_LKE0_BABA_0002_0272874.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00076590.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00064440.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0044_00033144.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0002_00114247.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC68_KSJ0_BABA_0001_0093515.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0056982.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00011988.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00076855.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_0060854.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00016714.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00066069.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00019900.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0022_00082712.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0004_00030239.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00030402.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0011219.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0014_0011179.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00105693.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0008_00045881.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00009142.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0006_00027208.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00067442.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00076010.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0019_00014270.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00083223.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_0146761.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00168522.json': ['Graspers (straight) (OLYMPUS) (GRSL-UCOL)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0007_00057695.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0012_00022062.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00035783.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0036_00166687.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_00073872.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0008_0043347.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0045524.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC16_PJW0_LAR0_0008_0019867.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0016_00043388.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0008_0022045.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0040_00311420.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC16_KSH0_LDG0_0040_00292627.json': ['Needle Holder (AESCULAP) (NDHL-UCAE)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00131859.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00021368.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0032_00037594.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00114881.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0003_0037291.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0011_0022458.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00062834.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0002_00029023.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'GGHB_DC68_LJH0_BABA_0003_0015321.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0019638.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_00062030.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC16_KMJ0_RHC0_0001_00067015.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0020_00059060.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0007_00031423.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_0047802.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)',\n",
       "  'Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00068900.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0011_00069590.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_00091854.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0020_0008532.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0021_0106991.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_00019332.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0000_0078510.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0044341.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0009_00038701.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'GGHB_DC68_LJH0_BABA_0002_0006061.json': ['Prograsp (da Vinci) (GRSR-UCDV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0001_00058536.json': ['Polymer Clip Applier (CLAL-UCUK)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0020_0008432.json': ['Suction Irrigator (SIRL-UCUK)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0008_00082396.json': ['Metal Clip Applier (Medtronic) (CLAL-M3MT)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0030021.json': ['R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RALP_0031_0058571.json': ['R-Scissors (da Vinci) (MSCR-UCDV)',\n",
       "  'R-Maryland (da Vinci) (BFCR-MADV)'],\n",
       " 'SNUH_DC07_JCW0_RLPN_0002_0123309.json': ['R-LND (da Vinci) (NDDR-LADV)'],\n",
       " 'SNUH_DC08_KHS0_ENME_0039_00061388.json': ['R-Scissors (da Vinci) (MSCR-UCDV)'],\n",
       " ...}"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# all_dic 은 key 값은 사진, value값은 수술도구인 딕셔너리.\n",
    "# 만들기 어려웠음! #수고했음!\n",
    "# 이렇게 하면 사진 하나에서 여러 어노테이션을 동시에 볼 수 있음.\n",
    "\n",
    "all_dic = {}\n",
    "\n",
    "for files in all_data_in:\n",
    "    for i in data_dic:\n",
    "        if(files in data_dic[i]):\n",
    "            if(files in all_dic):\n",
    "                temp = all_dic[files]\n",
    "                temp.append(i)\n",
    "                all_dic[files] = temp\n",
    "            else:\n",
    "                all_dic[files] = [i]\n",
    "all_dic"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "from PIL import Image, ImageDraw\n",
    "\n",
    "def getData(file, number):\n",
    "    json_path = train_data + '/' + class_list[number] + '/' + file\n",
    "    json_data = getjson(json_path)\n",
    "    width = json_data['images']['width']\n",
    "    height = json_data['images']['height']\n",
    "    \n",
    "    image_path = train_data+'/'+class_list[number]+'/'+ file.rpartition('.')[0]+'.png'\n",
    "    mask = np.load('data/'+file.rpartition('.')[0]+'.npy') == (number + 1)\n",
    "    mask = np.expand_dims(mask.astype(int), -1)\n",
    "    '''\n",
    "    mask = np.zeros((height,width,1))\n",
    "    \n",
    "    json_path = train_data + '/' + class_list[number] + '/' + file\n",
    "    json_data = getjson(json_path)\n",
    "    for classes in class_list:\n",
    "        for json_seg in json_data['annotations']:\n",
    "            if(classes == json_seg['category_name']):\n",
    "                points = json_seg[\"points\"].copy()\n",
    "                points.append(points[0])\n",
    "                points = [(x[0],x[1]) for x in points]\n",
    "                img = Image.new('L', (width, height), 0)\n",
    "                ImageDraw.Draw(img).polygon(points, outline=1, fill=1)\n",
    "                mask[:,:,0] = np.array(img)\n",
    "    \n",
    "    #np.save('data/'+file.rpartition('.')[0], mask)\n",
    "    '''\n",
    "    \n",
    "    return image_path, mask"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def _process_pathnames(filename, number):\n",
    "    # We map this function onto each pathname pair\n",
    "    fname, labels = getData(filename, number)\n",
    "    \n",
    "    img_str = tf.io.read_file(fname)\n",
    "    img = tf.image.decode_png(img_str, channels=3) # RGB\n",
    "    \n",
    "    label_img = tf.convert_to_tensor(labels)\n",
    "    width = tf.cast(label_img.shape[0], dtype=tf.uint16)\n",
    "    height = tf.cast(label_img.shape[1], dtype=tf.uint16)\n",
    "    \n",
    "    resize = [288, 512]\n",
    "    img = tf.image.resize(img, resize)\n",
    "    label_img = tf.image.resize(label_img, resize)\n",
    "    \n",
    "#    label_cls = [int(tf.cast(label_img==i, dtype=tf.uint8).numpy().sum()>0) for i in range(1,10)]\n",
    "#    label_cls = tf.cast(tf.convert_to_tensor(label_cls), dtype=tf.float32)\n",
    "    \n",
    "#    label_img = [tf.cast(label_img==i, dtype=tf.uint8).numpy() for i in range(1,10)]\n",
    "#    label_img = tf.concat(label_img,axis=-1)\n",
    "    \n",
    "    label_cls = [1] if (tf.reduce_sum(label_img,[0,1])>0).numpy() else [0]\n",
    "    label_cls = tf.cast(label_cls, dtype=tf.float32)\n",
    "    \n",
    "    scale = 1 / 255.                \n",
    "    img = tf.cast(img, dtype=tf.float32) * scale\n",
    "    filename = tf.cast(filename, dtype=tf.string)\n",
    "    \n",
    "    return filename, width, height, img, label_img, label_cls"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "val_ratio = 0.2\n",
    "seed = 1234\n",
    "\n",
    "train_filenames = [None]*9\n",
    "test_filenames = [None]*9\n",
    "\n",
    "for i in range(9):\n",
    "    train_filenames[i], test_filenames[i] = train_test_split(all_data[i],\n",
    "                                                   test_size=val_ratio,\n",
    "                                                   random_state=seed,\n",
    "                                                  )\n",
    "\n",
    "#num_train_examples = len(train_filenames)\n",
    "#num_test_examples = len(test_filenames)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import tensorflow_addons as tfa\n",
    "\n",
    "img_shape = (288, 512, 3)\n",
    "    \n",
    "def shift_img(output_img, label_img, width_shift_range, height_shift_range):\n",
    "    \"\"\"This fn will perform the horizontal or vertical shift\"\"\"\n",
    "    if width_shift_range or height_shift_range:\n",
    "        if width_shift_range:\n",
    "            width_shift_range = tf.random.uniform([],  \n",
    "                                                  -width_shift_range * img_shape[1],\n",
    "                                                  width_shift_range * img_shape[1])\n",
    "        if height_shift_range:\n",
    "            height_shift_range = tf.random.uniform([],\n",
    "                                                   -height_shift_range * img_shape[0],\n",
    "                                                   height_shift_range * img_shape[0])\n",
    "        output_img = tfa.image.translate(output_img,\n",
    "                                         [width_shift_range, height_shift_range])\n",
    "        label_img = tfa.image.translate(label_img,\n",
    "                                        [width_shift_range, height_shift_range])\n",
    "    return output_img, label_img\n",
    "\n",
    "def flip_img(horizontal_flip, vertically_flip, tr_img, label_img):\n",
    "    if horizontal_flip:\n",
    "        flip_prob = tf.random.uniform([], 0.0, 1.0)\n",
    "        tr_img, label_img = tf.cond(tf.less(flip_prob, 0.5),\n",
    "                                lambda: (tf.image.flip_left_right(tr_img), tf.image.flip_left_right(label_img)), \n",
    "                                lambda: (tr_img, label_img))\n",
    "    if vertically_flip:\n",
    "        flip_prob = tf.random.uniform([], 0.0, 1.0)\n",
    "        tr_img, label_img = tf.cond(tf.less(flip_prob, 0.5),\n",
    "                                lambda: (tf.image.flip_up_down(tr_img), tf.image.flip_up_down(label_img)), \n",
    "                                lambda: (tr_img, label_img))\n",
    "    return tr_img, label_img\n",
    "\n",
    "\n",
    "def rotate_img(tr_img, label_img):\n",
    "    rotate_prob = tf.random.uniform([], 0.0, 1.0)\n",
    "    degrees = tf.random.uniform([], -45, 45)\n",
    "    tr_img, label_img = tf.cond(tf.less(rotate_prob, 0.5),\n",
    "                            lambda: (tfa.image.rotate(tr_img, degrees * math.pi / 180, interpolation = 'BILINEAR', fill_mode= 'reflect'), tfa.image.rotate(label_img, degrees * math.pi / 180, interpolation = 'BILINEAR' , fill_mode= 'reflect')), \n",
    "                            lambda: (tr_img, label_img))\n",
    "    return tr_img, label_img\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def _augment(filename, width, height, img, label_img, label_cls,\n",
    "             resize=None,    # Resize the image to some size e.g. [256, 256]\n",
    "             scale=1,    # Scale image e.g. 1 / 255.\n",
    "             hue_delta=0.01,\n",
    "             horizontal_flip=True,    # Random left right flip,\n",
    "             vertically_flip=True,    # Random up down flip,\n",
    "             width_shift_range=.1,    # Randomly translate the image horizontally\n",
    "             height_shift_range=.1):    # Randomly translate the image vertically\n",
    "    \n",
    "    if resize is not None:\n",
    "        # Resize both images\n",
    "        label_img = tf.image.resize(label_img, resize)\n",
    "        img = tf.image.resize(img, resize)\n",
    "    \n",
    "    if hue_delta:\n",
    "        img = tf.image.random_hue(img, hue_delta)\n",
    "    \n",
    "    img, label_img = flip_img(horizontal_flip, vertically_flip, img, label_img)                               \n",
    "    img, label_img = shift_img(img, label_img, width_shift_range, height_shift_range)                      \n",
    "    img, label_img = rotate_img(img, label_img)\n",
    "    label_img = tf.cast(label_img, dtype=tf.float32) * scale\n",
    "    img = tf.cast(img, dtype=tf.float32) * scale\n",
    "    return filename, width, height, img, label_img, label_cls"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "import functools\n",
    "        \n",
    "\n",
    "batch_size = 4\n",
    "max_epochs = 20\n",
    "        \n",
    "def gen_train_1():\n",
    "    for i in train_filenames[0]:\n",
    "        yield _process_pathnames(i,0)\n",
    "        \n",
    "def gen_train_2():\n",
    "    for i in train_filenames[1]:\n",
    "        yield _process_pathnames(i,1)\n",
    "                \n",
    "def gen_train_3():\n",
    "    for i in train_filenames[2]:\n",
    "        yield _process_pathnames(i,2)\n",
    "                \n",
    "def gen_train_4():\n",
    "    for i in train_filenames[3]:\n",
    "        yield _process_pathnames(i,3)\n",
    "                \n",
    "def gen_train_5():\n",
    "    for i in train_filenames[4]:\n",
    "        yield _process_pathnames(i,4)\n",
    "                \n",
    "def gen_train_6():\n",
    "    for i in train_filenames[5]:\n",
    "        yield _process_pathnames(i,5)\n",
    "                \n",
    "def gen_train_7():\n",
    "    for i in train_filenames[6]:\n",
    "        yield _process_pathnames(i,6)\n",
    "             \n",
    "def gen_train_8():\n",
    "    for i in train_filenames[7]:\n",
    "        yield _process_pathnames(i,7)\n",
    "                \n",
    "def gen_train_9():\n",
    "    for i in train_filenames[8]:\n",
    "        yield _process_pathnames(i,8)\n",
    "        \n",
    "    \n",
    "def gen_test_1():\n",
    "    for i in test_filenames[0]:\n",
    "        yield _process_pathnames(i,0)\n",
    "    \n",
    "def gen_test_2():\n",
    "    for i in test_filenames[1]:\n",
    "        yield _process_pathnames(i,1)\n",
    "    \n",
    "def gen_test_3():\n",
    "    for i in test_filenames[2]:\n",
    "        yield _process_pathnames(i,2)\n",
    "    \n",
    "def gen_test_4():\n",
    "    for i in test_filenames[3]:\n",
    "        yield _process_pathnames(i,3)\n",
    "    \n",
    "def gen_test_5():\n",
    "    for i in test_filenames[4]:\n",
    "        yield _process_pathnames(i,4)\n",
    "    \n",
    "def gen_test_6():\n",
    "    for i in test_filenames[5]:\n",
    "        yield _process_pathnames(i,5)\n",
    "    \n",
    "def gen_test_7():\n",
    "    for i in test_filenames[6]:\n",
    "        yield _process_pathnames(i,6)\n",
    "    \n",
    "def gen_test_8():\n",
    "    for i in test_filenames[7]:\n",
    "        yield _process_pathnames(i,7)\n",
    "    \n",
    "def gen_test_9():\n",
    "    for i in test_filenames[8]:\n",
    "        yield _process_pathnames(i,8)\n",
    "        \n",
    "    \n",
    "def get_dataset(preproc_fn = functools.partial(_augment),\n",
    "                 threads = 6,\n",
    "                 batch_size = batch_size,\n",
    "                 is_train = True):\n",
    "    \n",
    "    dataset = [None] * 9\n",
    "    \n",
    "    if is_train:\n",
    "        dataset[0] = tf.data.Dataset.from_generator(gen_train_1, (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32), ((), (), (), (288,512,3), (288,512,1), (1)))\n",
    "        dataset[1] = tf.data.Dataset.from_generator(gen_train_2, (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32), ((), (), (), (288,512,3), (288,512,1), (1)))\n",
    "        dataset[2] = tf.data.Dataset.from_generator(gen_train_3, (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32), ((), (), (), (288,512,3), (288,512,1), (1)))\n",
    "        dataset[3] = tf.data.Dataset.from_generator(gen_train_4, (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32), ((), (), (), (288,512,3), (288,512,1), (1)))\n",
    "        dataset[4] = tf.data.Dataset.from_generator(gen_train_5, (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32), ((), (), (), (288,512,3), (288,512,1), (1)))\n",
    "        dataset[5] = tf.data.Dataset.from_generator(gen_train_6, (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32), ((), (), (), (288,512,3), (288,512,1), (1)))\n",
    "        dataset[6] = tf.data.Dataset.from_generator(gen_train_7, (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32), ((), (), (), (288,512,3), (288,512,1), (1)))\n",
    "        dataset[7] = tf.data.Dataset.from_generator(gen_train_8, (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32), ((), (), (), (288,512,3), (288,512,1), (1)))\n",
    "        dataset[8] = tf.data.Dataset.from_generator(gen_train_9, (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32), ((), (), (), (288,512,3), (288,512,1), (1)))\n",
    "\n",
    "    else:\n",
    "        dataset[0] = tf.data.Dataset.from_generator(gen_test_1, (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32), ((), (), (), (288,512,3), (288,512,1), (1)))\n",
    "        dataset[1] = tf.data.Dataset.from_generator(gen_test_2, (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32), ((), (), (), (288,512,3), (288,512,1), (1)))\n",
    "        dataset[2] = tf.data.Dataset.from_generator(gen_test_3, (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32), ((), (), (), (288,512,3), (288,512,1), (1)))\n",
    "        dataset[3] = tf.data.Dataset.from_generator(gen_test_4, (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32), ((), (), (), (288,512,3), (288,512,1), (1)))\n",
    "        dataset[4] = tf.data.Dataset.from_generator(gen_test_5, (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32), ((), (), (), (288,512,3), (288,512,1), (1)))\n",
    "        dataset[5] = tf.data.Dataset.from_generator(gen_test_6, (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32), ((), (), (), (288,512,3), (288,512,1), (1)))\n",
    "        dataset[6] = tf.data.Dataset.from_generator(gen_test_7, (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32), ((), (), (), (288,512,3), (288,512,1), (1)))\n",
    "        dataset[7] = tf.data.Dataset.from_generator(gen_test_8, (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32), ((), (), (), (288,512,3), (288,512,1), (1)))\n",
    "        dataset[8] = tf.data.Dataset.from_generator(gen_test_9, (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32), ((), (), (), (288,512,3), (288,512,1), (1)))\n",
    "    \n",
    "#    if is_train:\n",
    "#        for i in range(9):\n",
    "#            dataset[i] = dataset[i].map(preproc_fn, num_parallel_calls=threads)\n",
    "    for i in range(9):\n",
    "        dataset[i] = dataset[i].batch(batch_size)\n",
    "        \n",
    "    return dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_dataset = get_dataset(is_train=True)\n",
    "test_dataset = get_dataset(is_train=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[<BatchDataset shapes: ((None,), (None,), (None,), (None, 288, 512, 3), (None, 288, 512, 1), (None, 1)), types: (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32)>, <BatchDataset shapes: ((None,), (None,), (None,), (None, 288, 512, 3), (None, 288, 512, 1), (None, 1)), types: (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32)>, <BatchDataset shapes: ((None,), (None,), (None,), (None, 288, 512, 3), (None, 288, 512, 1), (None, 1)), types: (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32)>, <BatchDataset shapes: ((None,), (None,), (None,), (None, 288, 512, 3), (None, 288, 512, 1), (None, 1)), types: (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32)>, <BatchDataset shapes: ((None,), (None,), (None,), (None, 288, 512, 3), (None, 288, 512, 1), (None, 1)), types: (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32)>, <BatchDataset shapes: ((None,), (None,), (None,), (None, 288, 512, 3), (None, 288, 512, 1), (None, 1)), types: (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32)>, <BatchDataset shapes: ((None,), (None,), (None,), (None, 288, 512, 3), (None, 288, 512, 1), (None, 1)), types: (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32)>, <BatchDataset shapes: ((None,), (None,), (None,), (None, 288, 512, 3), (None, 288, 512, 1), (None, 1)), types: (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32)>, <BatchDataset shapes: ((None,), (None,), (None,), (None, 288, 512, 3), (None, 288, 512, 1), (None, 1)), types: (tf.string, tf.uint16, tf.uint16, tf.float32, tf.float32, tf.float32)>]\n"
     ]
    }
   ],
   "source": [
    "print(train_dataset)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras import layers\n",
    "from tensorflow.keras import models\n",
    "from tensorflow.keras import losses"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Conv(tf.keras.layers.Layer):\n",
    "    def __init__(self, num_filters, kernel_size):\n",
    "        super(Conv, self).__init__()\n",
    "        self.conv = layers.Conv2D(num_filters, kernel_size, padding='same')\n",
    "        self.bn = layers.BatchNormalization()\n",
    "        #self.dropout = layers.Dropout(0.2)\n",
    "\n",
    "    def call(self, inputs):\n",
    "        x = self.conv(inputs)\n",
    "        x = self.bn(x)\n",
    "        #x = self.dropout(x)\n",
    "        x = layers.ReLU()(x)\n",
    "\n",
    "        return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "class ConvBlock(tf.keras.layers.Layer):\n",
    "    def __init__(self, num_filters):\n",
    "        super(ConvBlock, self).__init__()\n",
    "        self.conv1 = Conv(num_filters, 3)\n",
    "        self.conv2 = Conv(num_filters, 3)\n",
    "\n",
    "    def call(self, inputs):\n",
    "        conv_block = self.conv1(inputs)\n",
    "        conv_block = self.conv2(conv_block)\n",
    "\n",
    "        return conv_block\n",
    "\n",
    "class EncoderBlock(tf.keras.layers.Layer):\n",
    "    def __init__(self, num_filters):\n",
    "        super(EncoderBlock, self).__init__()\n",
    "        self.conv_block = ConvBlock(num_filters)\n",
    "        self.encoder_pool = layers.Conv2D(num_filters, 3, padding='same', strides=2, activation='relu')\n",
    "        # layers.MaxPool2D((2,2))\n",
    "\n",
    "    def call(self, inputs):\n",
    "        encoder = self.conv_block(inputs)\n",
    "        encoder_pool = self.encoder_pool(encoder)\n",
    "\n",
    "        return encoder_pool, encoder                                         \n",
    "\n",
    "\n",
    "class DecoderBlock(tf.keras.layers.Layer):                                         \n",
    "    def __init__(self, num_filters):\n",
    "        super(DecoderBlock, self).__init__()\n",
    "        self.convT = layers.Conv2DTranspose(num_filters, 3, strides=2, padding='same')\n",
    "        self.bn = layers.BatchNormalization()\n",
    "        self.conv_block = ConvBlock(num_filters)\n",
    "\n",
    "    def call(self, input_tensor, concat_tensor):\n",
    "        decoder = self.convT(input_tensor)\n",
    "        decoder = self.bn(decoder)\n",
    "        decoder = layers.ReLU()(decoder)\n",
    "        decoder = layers.concatenate([decoder,concat_tensor]) # tf.concat([decoder, concat_tensor], -1)\n",
    "        decoder = self.conv_block(decoder)\n",
    "\n",
    "        return decoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "class UNet(tf.keras.Model):\n",
    "    def __init__(self):\n",
    "        super(UNet, self).__init__()\n",
    "        self.encoder_block1 = EncoderBlock(32)\n",
    "        self.encoder_block2 = EncoderBlock(64)\n",
    "        self.encoder_block3 = EncoderBlock(128)\n",
    "        self.encoder_block4 = EncoderBlock(256)\n",
    "        \n",
    "        self.center = ConvBlock(512)\n",
    "        \n",
    "        self.flatten = layers.Flatten()\n",
    "        self.output_linear = tf.keras.layers.Dense(1, activation='sigmoid')\n",
    "        \n",
    "        self.decoder_block4 = DecoderBlock(256)\n",
    "        self.decoder_block3 = DecoderBlock(128)\n",
    "        self.decoder_block2 = DecoderBlock(64) \n",
    "        self.decoder_block1 = DecoderBlock(32)  \n",
    "\n",
    "        self.output_conv = layers.Conv2D(1, 1, padding='same', activation='sigmoid')\n",
    "\n",
    "    def call(self, inputs):\n",
    "#        encoder0_pool, encoder0_skip = self.encoder_block0(inputs)\n",
    "        encoder1_pool, encoder1_skip = self.encoder_block1(inputs)\n",
    "        encoder2_pool, encoder2_skip = self.encoder_block2(encoder1_pool)\n",
    "        encoder3_pool, encoder3_skip = self.encoder_block3(encoder2_pool)\n",
    "        encoder4_pool, encoder4_skip = self.encoder_block4(encoder3_pool)\n",
    "        \n",
    "        center = self.center(encoder4_pool)\n",
    "        cls = self.output_linear(self.flatten(center))\n",
    "        \n",
    "        decoder4 = self.decoder_block4(center,encoder4_skip)\n",
    "        decoder3 = self.decoder_block3(decoder4,encoder3_skip)\n",
    "        decoder2 = self.decoder_block2(decoder3,encoder2_skip)\n",
    "        decoder1 = self.decoder_block1(decoder2,encoder1_skip)\n",
    "#        decoder0 = self.decoder_block0(decoder1, encoder0_skip)\n",
    "\n",
    "        outputs = self.output_conv(decoder1)\n",
    "        \n",
    "        return outputs, cls"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = [None] * 9\n",
    "for i in range(9):\n",
    "    model[i] = UNet()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(8, 288, 512, 1) (8, 1)\n",
      "Model: \"u_net\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "encoder_block (EncoderBlock) multiple                  19648     \n",
      "_________________________________________________________________\n",
      "encoder_block_1 (EncoderBloc multiple                  92864     \n",
      "_________________________________________________________________\n",
      "encoder_block_2 (EncoderBloc multiple                  370048    \n",
      "_________________________________________________________________\n",
      "encoder_block_3 (EncoderBloc multiple                  1477376   \n",
      "_________________________________________________________________\n",
      "conv_block_4 (ConvBlock)     multiple                  3544064   \n",
      "_________________________________________________________________\n",
      "flatten (Flatten)            multiple                  0         \n",
      "_________________________________________________________________\n",
      "dense (Dense)                multiple                  294913    \n",
      "_________________________________________________________________\n",
      "decoder_block (DecoderBlock) multiple                  2952960   \n",
      "_________________________________________________________________\n",
      "decoder_block_1 (DecoderBloc multiple                  739200    \n",
      "_________________________________________________________________\n",
      "decoder_block_2 (DecoderBloc multiple                  185280    \n",
      "_________________________________________________________________\n",
      "decoder_block_3 (DecoderBloc multiple                  46560     \n",
      "_________________________________________________________________\n",
      "conv2d_22 (Conv2D)           multiple                  33        \n",
      "=================================================================\n",
      "Total params: 9,722,946\n",
      "Trainable params: 9,716,098\n",
      "Non-trainable params: 6,848\n",
      "_________________________________________________________________\n",
      "(8, 288, 512, 1) (8, 1)\n",
      "Model: \"u_net_1\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "encoder_block_4 (EncoderBloc multiple                  19648     \n",
      "_________________________________________________________________\n",
      "encoder_block_5 (EncoderBloc multiple                  92864     \n",
      "_________________________________________________________________\n",
      "encoder_block_6 (EncoderBloc multiple                  370048    \n",
      "_________________________________________________________________\n",
      "encoder_block_7 (EncoderBloc multiple                  1477376   \n",
      "_________________________________________________________________\n",
      "conv_block_13 (ConvBlock)    multiple                  3544064   \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          multiple                  0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              multiple                  294913    \n",
      "_________________________________________________________________\n",
      "decoder_block_4 (DecoderBloc multiple                  2952960   \n",
      "_________________________________________________________________\n",
      "decoder_block_5 (DecoderBloc multiple                  739200    \n",
      "_________________________________________________________________\n",
      "decoder_block_6 (DecoderBloc multiple                  185280    \n",
      "_________________________________________________________________\n",
      "decoder_block_7 (DecoderBloc multiple                  46560     \n",
      "_________________________________________________________________\n",
      "conv2d_45 (Conv2D)           multiple                  33        \n",
      "=================================================================\n",
      "Total params: 9,722,946\n",
      "Trainable params: 9,716,098\n",
      "Non-trainable params: 6,848\n",
      "_________________________________________________________________\n",
      "(8, 288, 512, 1) (8, 1)\n",
      "Model: \"u_net_2\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "encoder_block_8 (EncoderBloc multiple                  19648     \n",
      "_________________________________________________________________\n",
      "encoder_block_9 (EncoderBloc multiple                  92864     \n",
      "_________________________________________________________________\n",
      "encoder_block_10 (EncoderBlo multiple                  370048    \n",
      "_________________________________________________________________\n",
      "encoder_block_11 (EncoderBlo multiple                  1477376   \n",
      "_________________________________________________________________\n",
      "conv_block_22 (ConvBlock)    multiple                  3544064   \n",
      "_________________________________________________________________\n",
      "flatten_2 (Flatten)          multiple                  0         \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              multiple                  294913    \n",
      "_________________________________________________________________\n",
      "decoder_block_8 (DecoderBloc multiple                  2952960   \n",
      "_________________________________________________________________\n",
      "decoder_block_9 (DecoderBloc multiple                  739200    \n",
      "_________________________________________________________________\n",
      "decoder_block_10 (DecoderBlo multiple                  185280    \n",
      "_________________________________________________________________\n",
      "decoder_block_11 (DecoderBlo multiple                  46560     \n",
      "_________________________________________________________________\n",
      "conv2d_68 (Conv2D)           multiple                  33        \n",
      "=================================================================\n",
      "Total params: 9,722,946\n",
      "Trainable params: 9,716,098\n",
      "Non-trainable params: 6,848\n",
      "_________________________________________________________________\n",
      "(8, 288, 512, 1) (8, 1)\n",
      "Model: \"u_net_3\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "encoder_block_12 (EncoderBlo multiple                  19648     \n",
      "_________________________________________________________________\n",
      "encoder_block_13 (EncoderBlo multiple                  92864     \n",
      "_________________________________________________________________\n",
      "encoder_block_14 (EncoderBlo multiple                  370048    \n",
      "_________________________________________________________________\n",
      "encoder_block_15 (EncoderBlo multiple                  1477376   \n",
      "_________________________________________________________________\n",
      "conv_block_31 (ConvBlock)    multiple                  3544064   \n",
      "_________________________________________________________________\n",
      "flatten_3 (Flatten)          multiple                  0         \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              multiple                  294913    \n",
      "_________________________________________________________________\n",
      "decoder_block_12 (DecoderBlo multiple                  2952960   \n",
      "_________________________________________________________________\n",
      "decoder_block_13 (DecoderBlo multiple                  739200    \n",
      "_________________________________________________________________\n",
      "decoder_block_14 (DecoderBlo multiple                  185280    \n",
      "_________________________________________________________________\n",
      "decoder_block_15 (DecoderBlo multiple                  46560     \n",
      "_________________________________________________________________\n",
      "conv2d_91 (Conv2D)           multiple                  33        \n",
      "=================================================================\n",
      "Total params: 9,722,946\n",
      "Trainable params: 9,716,098\n",
      "Non-trainable params: 6,848\n",
      "_________________________________________________________________\n",
      "(8, 288, 512, 1) (8, 1)\n",
      "Model: \"u_net_4\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "encoder_block_16 (EncoderBlo multiple                  19648     \n",
      "_________________________________________________________________\n",
      "encoder_block_17 (EncoderBlo multiple                  92864     \n",
      "_________________________________________________________________\n",
      "encoder_block_18 (EncoderBlo multiple                  370048    \n",
      "_________________________________________________________________\n",
      "encoder_block_19 (EncoderBlo multiple                  1477376   \n",
      "_________________________________________________________________\n",
      "conv_block_40 (ConvBlock)    multiple                  3544064   \n",
      "_________________________________________________________________\n",
      "flatten_4 (Flatten)          multiple                  0         \n",
      "_________________________________________________________________\n",
      "dense_4 (Dense)              multiple                  294913    \n",
      "_________________________________________________________________\n",
      "decoder_block_16 (DecoderBlo multiple                  2952960   \n",
      "_________________________________________________________________\n",
      "decoder_block_17 (DecoderBlo multiple                  739200    \n",
      "_________________________________________________________________\n",
      "decoder_block_18 (DecoderBlo multiple                  185280    \n",
      "_________________________________________________________________\n",
      "decoder_block_19 (DecoderBlo multiple                  46560     \n",
      "_________________________________________________________________\n",
      "conv2d_114 (Conv2D)          multiple                  33        \n",
      "=================================================================\n",
      "Total params: 9,722,946\n",
      "Trainable params: 9,716,098\n",
      "Non-trainable params: 6,848\n",
      "_________________________________________________________________\n",
      "(8, 288, 512, 1) (8, 1)\n",
      "Model: \"u_net_5\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "encoder_block_20 (EncoderBlo multiple                  19648     \n",
      "_________________________________________________________________\n",
      "encoder_block_21 (EncoderBlo multiple                  92864     \n",
      "_________________________________________________________________\n",
      "encoder_block_22 (EncoderBlo multiple                  370048    \n",
      "_________________________________________________________________\n",
      "encoder_block_23 (EncoderBlo multiple                  1477376   \n",
      "_________________________________________________________________\n",
      "conv_block_49 (ConvBlock)    multiple                  3544064   \n",
      "_________________________________________________________________\n",
      "flatten_5 (Flatten)          multiple                  0         \n",
      "_________________________________________________________________\n",
      "dense_5 (Dense)              multiple                  294913    \n",
      "_________________________________________________________________\n",
      "decoder_block_20 (DecoderBlo multiple                  2952960   \n",
      "_________________________________________________________________\n",
      "decoder_block_21 (DecoderBlo multiple                  739200    \n",
      "_________________________________________________________________\n",
      "decoder_block_22 (DecoderBlo multiple                  185280    \n",
      "_________________________________________________________________\n",
      "decoder_block_23 (DecoderBlo multiple                  46560     \n",
      "_________________________________________________________________\n",
      "conv2d_137 (Conv2D)          multiple                  33        \n",
      "=================================================================\n",
      "Total params: 9,722,946\n",
      "Trainable params: 9,716,098\n",
      "Non-trainable params: 6,848\n",
      "_________________________________________________________________\n",
      "(8, 288, 512, 1) (8, 1)\n",
      "Model: \"u_net_6\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "encoder_block_24 (EncoderBlo multiple                  19648     \n",
      "_________________________________________________________________\n",
      "encoder_block_25 (EncoderBlo multiple                  92864     \n",
      "_________________________________________________________________\n",
      "encoder_block_26 (EncoderBlo multiple                  370048    \n",
      "_________________________________________________________________\n",
      "encoder_block_27 (EncoderBlo multiple                  1477376   \n",
      "_________________________________________________________________\n",
      "conv_block_58 (ConvBlock)    multiple                  3544064   \n",
      "_________________________________________________________________\n",
      "flatten_6 (Flatten)          multiple                  0         \n",
      "_________________________________________________________________\n",
      "dense_6 (Dense)              multiple                  294913    \n",
      "_________________________________________________________________\n",
      "decoder_block_24 (DecoderBlo multiple                  2952960   \n",
      "_________________________________________________________________\n",
      "decoder_block_25 (DecoderBlo multiple                  739200    \n",
      "_________________________________________________________________\n",
      "decoder_block_26 (DecoderBlo multiple                  185280    \n",
      "_________________________________________________________________\n",
      "decoder_block_27 (DecoderBlo multiple                  46560     \n",
      "_________________________________________________________________\n",
      "conv2d_160 (Conv2D)          multiple                  33        \n",
      "=================================================================\n",
      "Total params: 9,722,946\n",
      "Trainable params: 9,716,098\n",
      "Non-trainable params: 6,848\n",
      "_________________________________________________________________\n",
      "(8, 288, 512, 1) (8, 1)\n",
      "Model: \"u_net_7\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "encoder_block_28 (EncoderBlo multiple                  19648     \n",
      "_________________________________________________________________\n",
      "encoder_block_29 (EncoderBlo multiple                  92864     \n",
      "_________________________________________________________________\n",
      "encoder_block_30 (EncoderBlo multiple                  370048    \n",
      "_________________________________________________________________\n",
      "encoder_block_31 (EncoderBlo multiple                  1477376   \n",
      "_________________________________________________________________\n",
      "conv_block_67 (ConvBlock)    multiple                  3544064   \n",
      "_________________________________________________________________\n",
      "flatten_7 (Flatten)          multiple                  0         \n",
      "_________________________________________________________________\n",
      "dense_7 (Dense)              multiple                  294913    \n",
      "_________________________________________________________________\n",
      "decoder_block_28 (DecoderBlo multiple                  2952960   \n",
      "_________________________________________________________________\n",
      "decoder_block_29 (DecoderBlo multiple                  739200    \n",
      "_________________________________________________________________\n",
      "decoder_block_30 (DecoderBlo multiple                  185280    \n",
      "_________________________________________________________________\n",
      "decoder_block_31 (DecoderBlo multiple                  46560     \n",
      "_________________________________________________________________\n",
      "conv2d_183 (Conv2D)          multiple                  33        \n",
      "=================================================================\n",
      "Total params: 9,722,946\n",
      "Trainable params: 9,716,098\n",
      "Non-trainable params: 6,848\n",
      "_________________________________________________________________\n",
      "(8, 288, 512, 1) (8, 1)\n",
      "Model: \"u_net_8\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "encoder_block_32 (EncoderBlo multiple                  19648     \n",
      "_________________________________________________________________\n",
      "encoder_block_33 (EncoderBlo multiple                  92864     \n",
      "_________________________________________________________________\n",
      "encoder_block_34 (EncoderBlo multiple                  370048    \n",
      "_________________________________________________________________\n",
      "encoder_block_35 (EncoderBlo multiple                  1477376   \n",
      "_________________________________________________________________\n",
      "conv_block_76 (ConvBlock)    multiple                  3544064   \n",
      "_________________________________________________________________\n",
      "flatten_8 (Flatten)          multiple                  0         \n",
      "_________________________________________________________________\n",
      "dense_8 (Dense)              multiple                  294913    \n",
      "_________________________________________________________________\n",
      "decoder_block_32 (DecoderBlo multiple                  2952960   \n",
      "_________________________________________________________________\n",
      "decoder_block_33 (DecoderBlo multiple                  739200    \n",
      "_________________________________________________________________\n",
      "decoder_block_34 (DecoderBlo multiple                  185280    \n",
      "_________________________________________________________________\n",
      "decoder_block_35 (DecoderBlo multiple                  46560     \n",
      "_________________________________________________________________\n",
      "conv2d_206 (Conv2D)          multiple                  33        \n",
      "=================================================================\n",
      "Total params: 9,722,946\n",
      "Trainable params: 9,716,098\n",
      "Non-trainable params: 6,848\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "for i in range(9):\n",
    "    outputs, cls = model[i](tf.random.normal([8, 288, 512, 3]))\n",
    "    print(outputs.shape, cls.shape)\n",
    "\n",
    "#print(predictions)\n",
    "    model[i].summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "def dice_coeff(y_true, y_pred): \n",
    "    smooth = 1e-10        \n",
    "    # Flatten\n",
    "    y_true_f = tf.reshape(y_true, [-1])\n",
    "    y_pred_f = tf.reshape(y_pred, [-1])\n",
    "    intersection = tf.reduce_sum(y_true_f * y_pred_f)\n",
    "    score = (2. * intersection + smooth) / (tf.reduce_sum(tf.square(y_true_f)) + \\\n",
    "                                            tf.reduce_sum(tf.square(y_pred_f)) + smooth)\n",
    "\n",
    "    return score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "def dice_loss(y_true, y_pred):\n",
    "    loss = 1 - dice_coeff(y_true, y_pred)\n",
    "    return loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "def bce_dice_loss(y_true, y_pred):\n",
    "    loss = tf.reduce_mean(losses.binary_crossentropy(y_true, y_pred)) + \\\n",
    "          dice_loss(y_true, y_pred)\n",
    "\n",
    "    return loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "bce_loss = tf.keras.losses.BinaryCrossentropy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import roc_curve\n",
    "from sklearn.metrics import auc\n",
    "\n",
    "def metric_function(y_true, y_true_cls, y_pred, y_pred_cls, num_classes=2):\n",
    "    # Flatten  \n",
    "    \n",
    "    y_true_cls = tf.cast(tf.round(y_true_cls), dtype=tf.int32)\n",
    "    y_pred_cls = tf.cast(tf.round(y_pred_cls), dtype=tf.int32)\n",
    "\n",
    "    y_true_cls = y_true_cls.numpy().reshape(-1)\n",
    "    y_pred_cls = y_pred_cls.numpy().reshape(-1)\n",
    "\n",
    "    y_true_f = tf.reshape(y_true, [-1])\n",
    "    y_pred_f = tf.reshape(y_pred, [-1])\n",
    "\n",
    "    y_true_f = tf.cast(tf.round(y_true_f), dtype=tf.int32).numpy()\n",
    "    y_pred_f = tf.cast(tf.round(y_pred_f), dtype=tf.int32).numpy()\n",
    "    \n",
    "    # calculate confusion matrix\n",
    "    labels = list(range(num_classes))\n",
    "    current = confusion_matrix(y_true_f, y_pred_f, labels=labels)\n",
    "    \n",
    "    # compute mean iou\n",
    "    intersection = np.diag(current)\n",
    "    ground_truth_set = current.sum(axis=1)\n",
    "    predicted_set = current.sum(axis=0)\n",
    "    union = ground_truth_set + predicted_set - intersection\n",
    "        \n",
    "    IoU = intersection / union.astype(np.float32)\n",
    "    \n",
    "    sensitivity = []\n",
    "    specificity = []\n",
    "    auc_metric = []\n",
    "    \n",
    "    # calculate roc curve\n",
    "    fpr, tpr, thresholds = roc_curve(y_true_cls, y_pred_cls)  \n",
    "    auc_metric.append(auc(fpr, tpr))\n",
    "\n",
    "    # calculate confusion matrix\n",
    "    tn, fp, fn, tp  = confusion_matrix(y_true_cls, y_pred_cls, labels=labels).ravel()\n",
    "    sensitivity.append(tp/(tp+fn+1e-6))#(tpr)\n",
    "    specificity.append(tn/(tn+fp+1e-6))#(1-fpr)\n",
    "    \n",
    "    sensitivity = np.mean(np.array(sensitivity))\n",
    "    specificity = np.mean(np.array(specificity))\n",
    "    auc_metric = np.mean(np.array(auc_metric))\n",
    "    \n",
    "    return sensitivity, specificity, auc_metric, np.mean(IoU)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "optimizer = [None] * 9\n",
    "checkpoint = [None] * 9\n",
    "\n",
    "checkpoint_prefix = [None] * 9 \n",
    "for i in range(9):\n",
    "    optimizer[i] = tf.keras.optimizers.Adam(1e-4)\n",
    "\n",
    "    checkpoint_prefix[i] = '/tf/backup/ckpt/U-NET/'+str(i)+'/'\n",
    "\n",
    "#    checkpoint[i].restore(tf.train.latest_checkpoint(checkpoint_prefix[i]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "GGHB_DC68_LJH0_BABA_0002_0018195.png\n",
      "cls : [1 0 0 0 0 0 0 0 0]\n",
      "pred : [0 0 0 0 0 0 0 0 1]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 400x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x600 with 9 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x600 with 9 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x600 with 9 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "metrics: [0.         0.84374997 0.421875   0.4904074  0.49041221]\n",
      "scores : 0.4629944386429399\n",
      "real_score : 0.4629973208586517\n",
      "Epoch Time : 2223.841061115265\n"
     ]
    },
    {
     "ename": "AttributeError",
     "evalue": "'list' object has no attribute 'save'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-34-4755e4d199d8>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m    132\u001b[0m \u001b[0;31m# saving (checkpoint) the model periodically\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    133\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mepoch\u001b[0m\u001b[0;34m+\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m%\u001b[0m \u001b[0msave_epochs\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 134\u001b[0;31m         \u001b[0mcheckpoint\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcheckpoint_prefix\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;31mAttributeError\u001b[0m: 'list' object has no attribute 'save'"
     ]
    }
   ],
   "source": [
    "from IPython.display import clear_output\n",
    "\n",
    "max_epochs = 50\n",
    "\n",
    "# save loss values for plot\n",
    "loss_history = [[None]] * 9\n",
    "metric_history = [[None]] * 9\n",
    "global_step = 0 # step 수 정의 (선택)\n",
    "print_count = 10 # tf.gradient_tape\n",
    "save_epochs = 2 # tf.gradient_tape\n",
    "#batch_step = (num_train_examples // batch_size) + (0 if num_train_examples % batch_size == 0 else 1)\n",
    "\n",
    "\n",
    "for epoch in range(max_epochs):\n",
    "    epoch_time = time.time()\n",
    "    for num in range(9):\n",
    "        print(num, end='. ')\n",
    "        for idx, (filename, width, height, images, labels, cls) in enumerate(train_dataset[num]): # 데이터 로드 파트\n",
    "\n",
    "            labels_no = tf.zeros((filename.shape[0],288,512,1))\n",
    "            cls_no = tf.expand_dims(tf.zeros(filename.shape[0]),-1)\n",
    "\n",
    "            start_time = time.time()\n",
    "            global_step = global_step + 1\n",
    "            \n",
    "            rand_int = (num + idx) % 9\n",
    "            \n",
    "            with tf.GradientTape(persistent=True) as tape: # 모델 학습 파트\n",
    "                predictions, pred_cls = model[num](images, training=True) # [batch_size, 256,256,3]\n",
    "                predictions_no, pred_cls_no = model[rand_int](images, training=True) # [batch_size, 256,256,3]\n",
    "                #predictions_no2, pred_cls_no2 = model[rand_int2](images, training=True) # [batch_size, 256,256,3]\n",
    "            \n",
    "                #predictions = tf.expand_dims(tf.argmax(predictions,axis=-1),-1)\n",
    "                loss = bce_dice_loss(labels, predictions) # label [batch_size, 256, 256, 1]\n",
    "                #loss = bce_dice_loss(labels, predictions * tf.reshape(tf.round(pred_cls),[pred_cls.shape[0],1,1,1]))\n",
    "                loss += bce_loss(cls, pred_cls)\n",
    "                \n",
    "                loss_no = bce_dice_loss(labels_no, predictions_no)\n",
    "                loss_no += bce_loss(cls_no, pred_cls_no)\n",
    "                \n",
    "                #loss_no2 = bce_dice_loss(labels_no, predictions_no2)\n",
    "                #loss_no2 += bce_loss(cls_no, pred_cls_no2)\n",
    "                \n",
    "            # 가중치 업데이트 파트\n",
    "            loss_history[num].append(loss)\n",
    "            gradients = tape.gradient(loss, model[num].trainable_variables)\n",
    "            gradients_no = tape.gradient(loss_no, model[rand_int].trainable_variables)\n",
    "            #gradients_no2 = tape.gradient(loss_no2, model[rand_int2].trainable_variables)\n",
    "\n",
    "            optimizer[num].apply_gradients(zip(gradients, model[num].trainable_variables))\n",
    "            optimizer[rand_int].apply_gradients(zip(gradients_no, model[rand_int].trainable_variables))\n",
    "            #optimizer[rand_int2].apply_gradients(zip(gradients_no2, model[rand_int2].trainable_variables))\n",
    "            del labels_no\n",
    "            del cls_no\n",
    "            \n",
    "            # 학습 상태 출력\n",
    "    epoch_duration = time.time() - epoch_time\n",
    "    print(\"Epoch Time :\", epoch_duration)\n",
    "    \n",
    "    num = np.random.randint(9)\n",
    "    mean = []\n",
    "    \n",
    "    # print sample image                     \n",
    "    for filename, width, height, test_images, test_label, test_cls in test_dataset[num].take(1):\n",
    "        clear_output(wait=True)\n",
    "        sample = 0\n",
    "        print(filename.numpy()[sample].decode('utf-8').rpartition('.')[0]+'.png')\n",
    "        \n",
    "        predictions = [None] * 9\n",
    "        pred_cls = [None] * 9\n",
    "        \n",
    "        for i in range(9):\n",
    "            predictions[i], pred_cls[i] = model[i](test_images, training=False)\n",
    "        \n",
    "        test_class = np.array([[0] * 9] * batch_size)\n",
    "        test_class[:,num] = tf.squeeze(test_cls)\n",
    "        \n",
    "        test_labels = np.zeros((batch_size, 288,512,9))\n",
    "        test_labels[:,:,:,num] = tf.squeeze(test_label)\n",
    "        \n",
    "        predictions = tf.concat(predictions,axis=-1)\n",
    "        pred_cls = tf.concat(pred_cls,axis=-1)\n",
    "        test_cls = np.array(test_class)\n",
    "        print('cls :', test_cls[sample])\n",
    "        print('pred :', pred_cls.numpy().round().astype(int)[sample])\n",
    "        \n",
    "        plt.figure(figsize=(4, 3))\n",
    "        plt.subplot(1, 1, 1)\n",
    "        plt.imshow(test_images[sample,:,:,:])\n",
    "        plt.title(\"Input image\")\n",
    "        plt.show()\n",
    "\n",
    "        plt.figure(figsize=(8, 6))\n",
    "        for i in range(9):\n",
    "            plt.subplot(3, 3, i+1)\n",
    "            plt.imshow(test_labels[sample,:,:,i])\n",
    "            plt.title(\"Actual Mask \"+str(i+1))\n",
    "        plt.show()\n",
    "\n",
    "        plt.figure(figsize=(8, 6))\n",
    "        for i in range(9):\n",
    "            plt.subplot(3, 3, i+1)\n",
    "            plt.imshow(predictions[sample,:,:,i])\n",
    "            plt.title(\"Predicted Mask \"+str(i+1))\n",
    "        plt.show()\n",
    "\n",
    "        plt.figure(figsize=(8, 6))\n",
    "        for i in range(9):\n",
    "            plt.subplot(3, 3, i+1)\n",
    "            plt.imshow(tf.round(predictions[sample,:,:,i]) * tf.round(pred_cls[sample,i]))\n",
    "            plt.title(\"Classification Mask \"+str(i+1))\n",
    "        plt.show()\n",
    "        \n",
    "        _, _, _, c_iou = metric_function(test_labels, test_cls, tf.round(predictions) * tf.reshape(tf.round(pred_cls),[pred_cls.shape[0],1,1,9]), pred_cls)\n",
    "\n",
    "        sen, spe, m_auc, iou = metric_function(test_labels, test_cls, predictions, pred_cls)\n",
    "        mean.append([sen, spe, m_auc, iou, c_iou])\n",
    "\n",
    "    mean = np.array(mean)\n",
    "    mean = np.mean(mean,axis=0)\n",
    "\n",
    "    metric_history[num].append([epoch, mean[0], mean[1], mean[2], mean[3], mean[4]])\n",
    "\n",
    "    score = mean[3]*0.6 + (mean[0]*0.1 + mean[1]*0.1 + mean[2]*0.2)\n",
    "    real_score = mean[4]*0.6 + (mean[0]*0.1 + mean[1]*0.1 + mean[2]*0.2)\n",
    "    print(\"metrics: {}\".format(mean))\n",
    "    print(\"scores : {}\".format(score))\n",
    "    print(\"real_score : {}\".format(real_score))\n",
    "    \n",
    "    print(\"Epoch Time :\", epoch_duration)\n",
    "    \n",
    "# saving (checkpoint) the model periodically\n",
    "    if (epoch+1) % save_epochs == 0:\n",
    "        checkpoint[i].save(checkpoint_prefix[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(9):\n",
    "    checkpoint[i] = tf.train.Checkpoint(optimizer=optimizer[i],\n",
    "                               model=model[i])\n",
    "    checkpoint[i].save(checkpoint_prefix[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "plt.figure(figsize=(8, 8))\n",
    "for i in range(9):\n",
    "    loss_history = np.asarray(loss_history[i])\n",
    "    plt.subplot(3,3,i+1)\n",
    "    plt.plot(loss_history[:,0], loss_history[:,1])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "metric_history = np.asarray(metric_history)\n",
    "plt.figure(figsize=(12, 8))\n",
    "plt.subplot(2, 3, 1)\n",
    "plt.plot(metric_history[:,0], metric_history[:,1])\n",
    "plt.title('sensitivity')\n",
    "\n",
    "plt.subplot(2, 3, 2)\n",
    "plt.plot(metric_history[:,0], metric_history[:,2])\n",
    "plt.title('specificity')\n",
    "\n",
    "plt.subplot(2, 3, 3)\n",
    "plt.plot(metric_history[:,0], metric_history[:,3])\n",
    "plt.title('AUC')\n",
    "\n",
    "plt.subplot(2, 3, 4)\n",
    "plt.plot(metric_history[:,0], metric_history[:,4])\n",
    "plt.title('iou')\n",
    "\n",
    "plt.subplot(2, 3, 5)\n",
    "plt.plot(metric_history[:,0], metric_history[:,5])\n",
    "plt.title('real_iou')\n",
    "\n",
    "plt.subplot(2, 3, 6)\n",
    "plt.plot(metric_history[:,0], metric_history[:,5]*0.6 + metric_history[:,1]*0.1 + metric_history[:,2]*0.1 + metric_history[:,3]*0.2)\n",
    "plt.title('real_score')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import cv2\n",
    "import pandas as pd\n",
    "\n",
    "mean = []\n",
    "\n",
    "# print sample image                     \n",
    "for dsf, (filename, width, height, images, labels, cls) in enumerate(test_dataset):\n",
    "    predictions, pred_cls = model(images, training=False)\n",
    "    \n",
    "    mask = np.zeros((predictions.shape[0],288,512,9))\n",
    "    \n",
    "    for sample in range(predictions.shape[0]):\n",
    "        \n",
    "        no_found = True\n",
    "        for idx, img_cls in enumerate(pred_cls.numpy().round()[sample]):\n",
    "            sample_label = predictions[sample].numpy().round()[:,:,idx]\n",
    "\n",
    "            if(img_cls==1 and sample_label.sum()>1500):\n",
    "                \n",
    "                img = (sample_label*255).astype(np.uint8)\n",
    "                \n",
    "                contours, hierarchy = cv2.findContours(img, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_NONE)\n",
    "                \n",
    "                maximum = 0\n",
    "                maximum_iter = 0\n",
    "                for i, cnt in enumerate(contours):\n",
    "                    if(maximum < cv2.contourArea(cnt)):\n",
    "                        maximum = cv2.contourArea(cnt)\n",
    "                        maximum_iter = i\n",
    "                points = contours[maximum_iter]\n",
    "\n",
    "                for i in range(100):\n",
    "                    pl = cv2.approxPolyDP(points, (i/100)*cv2.arcLength(points, True), True)\n",
    "                    if(pl.shape[0]<=10):\n",
    "                        break\n",
    "                pl = pl.squeeze()\n",
    "\n",
    "                for i in range(10-len(pl)):\n",
    "                    pl = np.concatenate((pl, pl[0].reshape(1,2)), axis=0)\n",
    "\n",
    "                pl = [(x[0],x[1]) for x in pl]\n",
    "                _img = Image.new('L', (512,288), 0)\n",
    "                ImageDraw.Draw(_img).polygon(pl, outline=1, fill=1)\n",
    "                mask[sample,:,:,idx] = np.array(_img)#tf.image.resize(np.expand_dims(np.array(_img), axis=-1),[288,512]).numpy().squeeze()\n",
    "\n",
    "    sen, spe, m_auc, iou = metric_function(labels, cls, mask, tf.round(pred_cls).numpy())\n",
    "    mean.append(sen*0.1 + spe*0.1 + m_auc*0.2 + iou*0.6)\n",
    "\n",
    "    print('score :',sen*0.1 + spe*0.1 + m_auc*0.2 + iou*0.6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#테스트 스코어\n",
    "np.mean(np.array(mean))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "inference_list = os.listdir(inference_data)\n",
    "inference_list.sort()\n",
    "inference_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def _process_inference(filename):\n",
    "    # We map this function onto each pathname pair\n",
    "    img_path = inference_data+'/'+filename\n",
    "    \n",
    "    img_str = tf.io.read_file(img_path)\n",
    "    img = tf.image.decode_png(img_str, channels=3) # RGB\n",
    "    \n",
    "    width = tf.cast(img.shape[0], dtype=tf.uint16)\n",
    "    height = tf.cast(img.shape[1], dtype=tf.uint16)\n",
    "    \n",
    "    resize = [288, 512]\n",
    "    img = tf.image.resize(img, resize)\n",
    "    \n",
    "    scale = 1 / 255.                \n",
    "    img = tf.cast(img, dtype=tf.float32) * scale\n",
    "    filename = tf.cast(filename, dtype=tf.string)\n",
    "    \n",
    "    return filename, width, height, img"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def gen_infer():\n",
    "    for i in inference_list:\n",
    "        yield _process_inference(i)\n",
    "    \n",
    "def dataset_infer(batch_size=4):\n",
    "    \n",
    "    dataset = tf.data.Dataset.from_generator(gen_infer, (tf.string, tf.uint16, tf.uint16, tf.float32), ((), (), (), (288,512,3)))\n",
    "    \n",
    "    dataset = dataset.batch(batch_size)\n",
    "    return dataset\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "infer_dataset = dataset_infer()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import cv2\n",
    "import pandas as pd\n",
    "\n",
    "results = pd.DataFrame(columns=['file', 'class', 'x1', 'y1', 'x2', 'y2', 'x3', 'y3', 'x4', 'y4', 'x5', 'y5', 'x6', 'y6', 'x7', 'y7', 'x8', 'y8', 'x9', 'y9', 'x10', 'y10'])\n",
    "count = 0\n",
    "\n",
    "# print sample image                     \n",
    "for filename, width, height, images in infer_dataset:\n",
    "    predictions, pred_cls = model(images, training=False)\n",
    "    for sample in range(predictions.shape[0]):\n",
    "        no_found = True\n",
    "        for idx, img_cls in enumerate(pred_cls.numpy().round()[sample]):\n",
    "            sample_label = predictions[sample].numpy().round()[:,:,idx]\n",
    "            if(img_cls==1 and sample_label.sum()>300):\n",
    "                img = tf.image.resize(np.expand_dims(sample_label, axis=-1), [width.numpy()[sample], height.numpy()[sample]])\n",
    "                img = img.numpy()\n",
    "                img = (img*255).astype(np.uint8)\n",
    "                \n",
    "                contours, hierarchy = cv2.findContours(img, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_NONE)\n",
    "                \n",
    "                maximum = 0\n",
    "                maximum_iter = 0\n",
    "                for i, cnt in enumerate(contours):\n",
    "                    if(maximum < cv2.contourArea(cnt)):\n",
    "                        maximum = cv2.contourArea(cnt)\n",
    "                        maximum_iter = i\n",
    "                points = contours[maximum_iter]\n",
    "\n",
    "                for i in range(100):\n",
    "                    pl = cv2.approxPolyDP(points, (i/100)*cv2.arcLength(points, True), True)\n",
    "                    if(pl.shape[0]<=10):\n",
    "                        break\n",
    "                pl = pl.squeeze()\n",
    "\n",
    "                for i in range(10-len(pl)):\n",
    "                    pl = np.concatenate((pl, pl[0].reshape(1,2)), axis=0)\n",
    "                pl = pl.flatten()\n",
    "                result = [filename.numpy()[sample].decode('utf-8').rpartition('.')[0], idx]\n",
    "                result.extend(pl)\n",
    "                \n",
    "                results.loc[count] = result\n",
    "                no_found = False\n",
    "                count += 1\n",
    "        if(no_found):\n",
    "            result = [filename.numpy()[sample].decode('utf-8').rpartition('.')[0]]\n",
    "            result.extend([None] * 21)\n",
    "            results.loc[count] = result\n",
    "            count += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "results.to_csv('results.csv',index = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}