Hi Martin,

Can you try to remove the lesion intensity constraint on the T2 data (i.e., by modifying your command with --lesion-mask-pattern 0 1 0)  for a couple of cases and let us know if it helps? 

If not, would you be able to share with us one case where it fails? 

FYI: you can save a lesion probability map directly by using the flag "--save-posteriors Lesions" so you don't need to try different thresholds.

Thanks,
Stefano


Subject: [Freesurfer] Large number of false negative MS lesions in SAMSEG
Date: Sun, 14 Aug 2022 20:54:50 +0200
From: Martin Kavec <martin.kavec@gmail.com>
Reply-To: Freesurfer support list <freesurfer@nmr.mgh.harvard.edu>
To: Freesurfer mailing list <freesurfer@nmr.mgh.harvard.edu>


        External Email - Use Caution        

Hello,

I am testing SAMSEG on my group of MS patients. Unfortunately, even when dropping threshold to 0.005 I have large number of false negative lesions left behind by SAMSEG (80% in about 75 cases I tried), which are clearly visible lesions. I ran/checked registration between inputs, which is typically very good to excellent. Images are all 3D acquisitions , mprage, flair, and t2 with ~1 mm3 isotropic voxel. Here is the command, which I ran:

run_samseg --input mprage_reg.nii.gz flair.nii.gz t2_reg.nii.gz --pallidum-separate --lesion --lesion-mask-pattern 0 1 1 --threshold 0.005 --output samseg_threshold_0.005

and example of the lesion in occipital white matter in sagittal and axial orientations.

image.png
image.png

I would very much appreciate any leads as to how to improve the segmentation.

Thanks,

Martin