External Email - Use Caution
Hi Douglas Greve, thanks for the tip.
I tried to use SAMSEG on several images of the ADNI database obtaining very different values of the BV/CSF index than those defined in the reference article [1], more precisely they differ by an order of magnitude. Since they also used FreeSurfer for the calculation of the volumes, could you tell me if I am neglecting some interesting values or considering something that should be omitted instead?
The calculation of the index (with the anatomical regions of interest) for the generic image with ID 1075136 is shown in the attachment.
[1] https://secure-web.cisco.com/1kgOmQDmyoM6VUUACvqFq9dlBcuf7-_VdI2lRjPn0c1Z4DB...
Best regards,
Michele Grasso
________________________________ Da: Michele GRASSO Inviato: martedì 12 luglio 2022 10:57 A: freesurfer@nmr.mgh.harvard.edu freesurfer@nmr.mgh.harvard.edu Oggetto: BV/CSF index and Sulcal CSF
Hello FreeSurfer Developers,
My name is Michele Grasso and I am a student of the master's degree course in Computer Engineering at the University of Catania (Italy).
I am interested in calculating the BV/CSF index [1] which is equal to the ratio between the sum of the volumes of gray matter and white matter and the volume of the CSF. Regarding the volume of gray and white matter, I referred to the TotalGrayVol and CerebralWhiteMatterVol volume statistics reported in the aseg.stats file. CSF volume was considered as the sum of two contributions: ventricular (Left Lateral Ventricle (SegId 4) + Left Inferior Lateral Ventricle (SegId 5) + Right lateral Ventricle (SegId 43) + Right Inf Lateral Ventricle (SegId 44) + 3rd Ventricle (SegId 14) + 4th Ventricle (SegId 15) + 5th Ventricle (SegId 72) + Right choroids plexus (SegId 63) + Left choroids plexus (SegId 31)) and sulcal CSF. I searched the list and the sulcal CSF is not segmented by Freesurfer as it is indistinguishable from bone on a T1-weighted MRI.
What parameters in the aseg.stats file to consider to have a more likely measurement of the sulcal CSF? Could eTIV-BrainSegVolNotVent-ventricular volume be a good approximation? Finally, do I have to take into account the CSF statistic reported in the aforementioned file or ignore it completely?
Thank you,
Michele Grasso
[1] https://secure-web.cisco.com/1kgOmQDmyoM6VUUACvqFq9dlBcuf7-_VdI2lRjPn0c1Z4DB...
