Dear FreeSurfer-User,
some of my questions are mentioned in some older postings, but I still struggle with some basic questions. We are planning to use the new hippocampal subfield segmentation of FS 5.1 with 3 T Siemens TRIO with 32 channel head coil. Still, I do not fully understand how to optimize our image processeing pipeline:
1. Freesurfer works internally with 1x1x1 mm3 but the hippocampal segmenat uses a higher res. of 0.5 x 0.5 x 0.5 mm3. So should we instead use a higher resolution for the hippocampal subfield seg.? 2. I understand that the recent FS version is not working well with high resolution images, e.g. the atlases are not provided in a higher res, however the hippocampal subfield segmentation seems to make use of aseg.stats. So is is possible at all to use the FS-pipleine in this case (hippocampal subfields)? 3. Instead of using a high resolution image, we could average over several pictures. When I tried this, I had the visual impression, that there was almost no gain in this, the volume estimates were bouncing within a range of 10% but did not converge when averaging from scan 1-5 to a specific value. So is it necessary for a good quality to average images when using hippocampal subfield segmentation ?
Thanks for your efforts!
Best,
Dr. med. Oliver Grimm Zentralinstitut für Seelische Gesundheit
www.zi-mannheim.de
SNiP | Systems Neuroscience in Psychiatry
68159 Mannheim
Germany
Tel: +49-621-1703-6341 Fax: +49-621-1703-6305 email oliver.grimm@zi-mannheim.de mailto:oliver.grimm@zi-mannheim.de