We are using SCT (v4.1.1) to detect lesions on MRI images of our patients. We came across an issue, where in T2-Stir image, we get 7 lesions even though there are none according to the doctors.
Any idea how I can improve the detection of lesions in MRI images that are incorrectly recognised?
Thank you for your feedback. Usually radiologists cross-validate the presence/absence of lesions using multi-contrasts. Currently,
sct_deepseg_lesion only considers single contrasts, but we are actively working on improving the algorithm to be able to use multiple contrasts, thereby improving the sensitivity and specificity of lesion segmentation.