Both TotalSegmentator AI segmentation and pyRadiomics are now accessible through TriDFusion extension modules.
https://github.com/dicomtools/TriDFusion
TotalSegmentator AI segmentation:
The machine learning module can segment 104 different bones, organ, muscles automatically. TriDFusion (3DF) dialog menu allows the user to select the classes of organ structures the user wishes to segment. It also allows the user to save\ load a pre-defined organ set through the use of an interface protocol. TriDFusion (3DF) can copy the volumes-of-interest from one image series to another modality. The segmentation is saved as a DICOM RT-Structure or DICOM mask, meaning that it is stored as an object for later use.
https://github.com/wasserth/TotalSegmentator
Single click 3D SPECT lung shunt processing:
This tool automatically segments the liver and lungs on the CT and automatically copies the volume-of-interest to the SPECT AC. The counts within each respective volume of the SPECT are then used to estimate the lung shunt fraction (LSF) of the series. A report is automatically generated and can be exported as a .pdf or DICOM screen capture.
Single click 3D SPECT lung ratio processing:
This tool automatically segments the individual lung lobes of both the left and right lung on the CT. The segmentation is automatically copied on the SPECT AC. The ratio of the left and right lung and of each individual lobes is calculated from the counts and volume of the SPECT AC. A report is automatically generated and can be exported as a .pdf or DICOM screen capture.
Radiomics Analysis:
A new feature is available to perform an extraction of the radiomics using the pyRadiomics package. The TriDFusion (3DF) interface allow the user to select between the different feature class, setting and image type. The user can save\ load a defined organ set through the use of an interface protocol. The radiomics results can be extracted from either one single volume-of-interest, or all VOIs and the results exported into an .xls excel file.
The TriDFusion (3DF) radiomics module is a complement to the TriDFusion (3DF) semi-automatic PET tumor segmentation contour tool that extracts the main object statistical attributes that include Total, Mean, Min, Max, Median Deviation, Peak, Max Diameter cm, Max SAD cm, Area cm2 and Volume cm3. This data is collected for each region-of-interest in addition to the extracted MTV and TLG for all lesion volumes-of-interest in the PET scan.