Clinical Study with Automatic ROI Placement

Semi-automated workflow for annotating a clinical study, including spectral unmixing, tissue segmentation and automatic ROI placement in the target muscle.

Target user: A medical scientist conducting a large clinical OA study. Study subjects are annotated scan by scan. For each scan, a segmentation mask is generated by the segmentation network, the scan is spectrally unmixed, and an ROI template is automatically placed into the target tissue. The pixel intensities within this region are summarized using statistical features and collected across scans. Parameters should remain fixed throughout the analysis. Finally, the collected data is exported into a spreadsheet for subsequent analysis.

Related pages: Segmentation · Unmixing · Keyboard Shortcuts