External Email - Use Caution
Hello. I am running run_samseg_long on Freesurfer 8.1.0 with longitudinal MS T1W data and I am trying to resolve lesion undersegmentation. Whenever I lower --threshold manually (e.g., --threshold 0.2), it seems to ignore it, outputting "thresholding posterior of Lesions with threshold: 0.3". To try thresholding manually, I added to the command --save-posteriors and --save-probabilities, but I'm not exactly sure what the outputted "Lesions.mgz" files in the posteriors and probabilities folders represent (and why the "Lesions.mgz" file in the probabilities folder has three frames). Can I please get advice on how to compute lesion volumes and masks at a threshold that is lower than 0.3? Thank you!
freesurfer@nmr.mgh.harvard.edu