[WIP] [BUG] Fix reading bad channels in cnt files (mne-tools#12189)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Eric Larson <[email protected]>

doc/changes/devel.rst

@@ -95,6 +95,8 @@ Bugs
 - Fix bug with :func:`mne.io.read_raw_eeglab` and :func:`mne.read_epochs_eeglab` where automatic fiducial detection would fail for certain files (:gh:`12165` by `Clemens Brunner`_)
 - Fix concatenation of ``raws`` with ``np.nan`` in the device to head transformation (:gh:`12198` by `Mathieu Scheltienne`_)
 - Fix bug with :func:`mne.viz.plot_compare_evokeds` where the title was not displayed when ``axes='topo'`` (:gh:`12192` by `Jacob Woessner`_)
+- Fix bug with :func:`mne.io.read_raw_cnt` where the bad channels were not properly read (:gh:`12189` by `Jacob Woessner`_)
+
 
 API changes
 ~~~~~~~~~~~

mne/io/cnt/cnt.py

@@ -14,7 +14,7 @@
 from ..._fiff.utils import _create_chs, _find_channels, _mult_cal_one, read_str
 from ...annotations import Annotations
 from ...channels.layout import _topo_to_sphere
-from ...utils import _check_option, fill_doc, warn
+from ...utils import _check_option, _validate_type, fill_doc, warn
 from ..base import BaseRaw
 from ._utils import (
     CNTEventType3,
@@ -169,6 +169,8 @@ def read_raw_cnt(
     emg=(),
     data_format="auto",
     date_format="mm/dd/yy",
+    *,
+    header="auto",
     preload=False,
     verbose=None,
 ):
@@ -219,6 +221,13 @@ def read_raw_cnt(
         Defaults to ``'auto'``.
     date_format : ``'mm/dd/yy'`` | ``'dd/mm/yy'``
         Format of date in the header. Defaults to ``'mm/dd/yy'``.
+    header : ``'auto'`` | ``'new'`` | ``'old'``
+        Defines the header format. Used to describe how bad channels
+        are formatted. If auto, reads using old and new header and
+        if either contain a bad channel make channel bad.
+        Defaults to ``'auto'``.
+
+        .. versionadded:: 1.6
     %(preload)s
     %(verbose)s
 
@@ -244,12 +253,13 @@ def read_raw_cnt(
         emg=emg,
         data_format=data_format,
         date_format=date_format,
+        header=header,
         preload=preload,
         verbose=verbose,
     )
 
 
-def _get_cnt_info(input_fname, eog, ecg, emg, misc, data_format, date_format):
+def _get_cnt_info(input_fname, eog, ecg, emg, misc, data_format, date_format, header):
     """Read the cnt header."""
     data_offset = 900  # Size of the 'SETUP' header.
     cnt_info = dict()
@@ -340,13 +350,23 @@ def _get_cnt_info(input_fname, eog, ecg, emg, misc, data_format, date_format):
         ch_names, cals, baselines, chs, pos = (list(), list(), list(), list(), list())
 
         bads = list()
+        _validate_type(header, str, "header")
+        _check_option("header", header, ("auto", "new", "old"))
         for ch_idx in range(n_channels):  # ELECTLOC fields
             fid.seek(data_offset + 75 * ch_idx)
             ch_name = read_str(fid, 10)
             ch_names.append(ch_name)
-            fid.seek(data_offset + 75 * ch_idx + 4)
-            if np.fromfile(fid, dtype="u1", count=1).item():
-                bads.append(ch_name)
+
+            # Some files have bad channels marked differently in the header.
+            if header in ("new", "auto"):
+                fid.seek(data_offset + 75 * ch_idx + 14)
+                if np.fromfile(fid, dtype="u1", count=1).item():
+                    bads.append(ch_name)
+            if header in ("old", "auto"):
+                fid.seek(data_offset + 75 * ch_idx + 4)
+                if np.fromfile(fid, dtype="u1", count=1).item():
+                    bads.append(ch_name)
+
             fid.seek(data_offset + 75 * ch_idx + 19)
             xy = np.fromfile(fid, dtype="f4", count=2)
             xy[1] *= -1  # invert y-axis
@@ -451,6 +471,11 @@ class RawCNT(BaseRaw):
         Defaults to ``'auto'``.
     date_format : ``'mm/dd/yy'`` | ``'dd/mm/yy'``
         Format of date in the header. Defaults to ``'mm/dd/yy'``.
+    header : ``'auto'`` | ``'new'`` | ``'old'``
+        Defines the header format. Used to describe how bad channels
+        are formatted. If auto, reads using old and new header and
+        if either contain a bad channel make channel bad.
+        Defaults to ``'auto'``.
     %(preload)s
     stim_channel : bool | None
         Add a stim channel from the events. Defaults to None to trigger a
@@ -478,6 +503,8 @@ def __init__(
         emg=(),
         data_format="auto",
         date_format="mm/dd/yy",
+        *,
+        header="auto",
         preload=False,
         verbose=None,
     ):  # noqa: D102
@@ -489,7 +516,7 @@ def __init__(
 
         input_fname = path.abspath(input_fname)
         info, cnt_info = _get_cnt_info(
-            input_fname, eog, ecg, emg, misc, data_format, _date_format
+            input_fname, eog, ecg, emg, misc, data_format, _date_format, header
         )
         last_samps = [cnt_info["n_samples"] - 1]
         super(RawCNT, self).__init__(

mne/io/cnt/tests/test_cnt.py

@@ -19,7 +19,7 @@
 
 
 @testing.requires_testing_data
-def test_data():
+def test_old_data():
     """Test reading raw cnt files."""
     with pytest.warns(RuntimeWarning, match="number of bytes"):
         raw = _test_raw_reader(
@@ -37,6 +37,39 @@ def test_data():
     assert raw.info["meas_date"] is None
 
 
+@testing.requires_testing_data
+def test_new_data():
+    """Test reading raw cnt files with different header."""
+    with pytest.warns(RuntimeWarning):
+        raw = read_raw_cnt(input_fname=fname_bad_spans, header="new")
+
+    assert raw.info["bads"] == ["F8"]  # test bads
+
+
+@testing.requires_testing_data
+def test_auto_data():
+    """Test reading raw cnt files with automatic header."""
+    with pytest.warns(RuntimeWarning):
+        raw = read_raw_cnt(input_fname=fname_bad_spans)
+
+    assert raw.info["bads"] == ["F8"]
+
+    with pytest.warns(RuntimeWarning, match="number of bytes"):
+        raw = _test_raw_reader(
+            read_raw_cnt, input_fname=fname, eog="auto", misc=["NA1", "LEFT_EAR"]
+        )
+
+    # make sure we use annotations event if we synthesized stim
+    assert len(raw.annotations) == 6
+
+    eog_chs = pick_types(raw.info, eog=True, exclude=[])
+    assert len(eog_chs) == 2  # test eog='auto'
+    assert raw.info["bads"] == ["LEFT_EAR", "VEOGR"]  # test bads
+
+    # the data has "05/10/200 17:35:31" so it is set to None
+    assert raw.info["meas_date"] is None
+
+
 @testing.requires_testing_data
 def test_compare_events_and_annotations():
     """Test comparing annotations and events."""