For any questions, recommendations, or feedback, please contact: [email protected] The upgraded XCAT phantom is available upon request from Dr. Paul Segars: [email protected] and additional information can be accessed at: https://olv.duke.edu/industry-investors/available-technologies/xcat/.
Summary:
• Novel lymphatic system was defined for the 4D-extended cardiac torso (XCAT) phantom
• A pipeline was developed to simulate and reconstruct PET images from the XCAT phantom, applying header information that is compatible with clinical radiology software
• As example application, simulated lymphoma patients were modelled using XCAT phantom and can be freely accessed
Reference:
Please use the following references if you publish results with help from this software tool:
R. Fedrigo, et al., “Development of scalable lymphatic system in the 4D XCAT phantom: Application to quantitative evaluation of lymphoma PET segmentations", Med. Phys., vol. 49, pp. 6871-6884, 2022.
The Matlab component of this framework is adapted from the PET simulation and image reconstruction tool, as described by:
S. Ashrafinia, et al., “Generalized PSF modeling for optimized quantitative-task performance”, Phys. Med. Biol., vol. 62, pp. 5149-5179, 2017.
Technical Description:
The novel lymphatic system for the 4D-extended cardiac torso (XCAT) phantom enhances the ability to model diseases, such as lymphoma. The lymphatic system (nodes, vessels) was defined using non-uniform rational basis spline (NURBS) surfaces. Multichannel large deformation diffeomorphic metric mapping (MC-LDDMM) method was used to propagate from the template phantom to different XCAT anatomies. The XCAT general parameter script was used to generate files that define the ground truth radioactivity and attenuation for a simulated patient.
Example files are provided, which can be used to model patients with non-Hodgkin’s lymphoma. Bulky tumours were created by altering lymph node morphology and function (nodes were expanded, stretched, converged, and had increased tracer uptake). This methodology is highly flexible and may be used in virtual clinical trials to simulate nodes with different pathology.
A framework was developed in Matlab and Python to simulate and reconstruct PET images using patients modelled with the XCAT phantom. Output files are converted to dicom and necessary header information is applied, such that the images can be viewed using clinical radiology software.
File Sequences:
- Please ensure that ground truth files generated from XCAT phantom are placed in the input folder.
- Run Main_PET_sim_recon from PET_sim_recon-master to perform PET simulation and reconstruction.
- Simulated images will appear in dicom-conversion folder.
- Run PT-bin-to-dicom to convert images to dicom and populate header information.
- Final simulated PET images are shown in output folder.
Compatibility note:
The MATLAB scripts were executed using the following toolboxes and versions.
MATLAB, version 9.8
Simulink 10.1
Computer Vision Toolbox 9.2
Curve Fitting Toolbox 3.5.11
DSP System Toolbox 9.10
Deep Learning Toolbox 14.0
Fuzzy Logic Toolbox 2.7
Image Processing Toolbox 11.1
Signal Processing Toolbox 8.4
Statistics and Machine Learning Toolbox 11.7
Symbolic Math Toolbox 8.5
Contact:
For questions, recommendations, or feedback, please contact: [email protected]