Hi Jo

the resolution requiirement is determined by cortical geometry. Once you get to about 2mm, you can no longer resolve the cortex in many places, and even worse you start losing the ability to resolve adjacent banks of a sulcus, resulting in topological catastrophes in the resulting models. The aseg probably works ok if that is all you are interested in, but you won't get good surfaces out of 7mm data

cheers Bruce

On Tue, 18 Sep 2018, 조성현 wrote:

External Email - Use Caution

Hello Freesurfer developers and subscribers, I'm attempting to run brain segmentation with low resolution mri(thickness ~= 7mm) for EEG analysis. I know that this is very low resolution to get good quality results. However, I think the low quality result is not meaningless in specific cases.

Therefore, I wonder if there is a tool to evaluate how low quality results are. In the 'freesurfer beginners guide', it says 'Thickness should not exceed 1.5mm'.  How the resolution requirement was determined? How did you know that the input of 1mm ^ 3 resolution is sufficient requirement? If I use 2mm^3 resolution input, what kind of result will I get?

It may be a little dumb, but I do not have an answer of this question. I would appreciate your reply.

Thanks, Jo