Hello,
I’m very new to neuroimaging and I have a few questions and need clarifications on how to preprocess MRI images
I currently have a set of 3T images with
dimensions: 176 x 512 x 512
voxel sizes: 1.000000, 0.468800, 0.468800
I’ve experimented with chucking the data immediately into freesurfer, and fastsurfer where the images were automatically resampled and isometricised to 1mm on freesurfer, and 1mm and 0.7mm on fastsrufer. The results looked reasonable but I was wondering if it would be better if I were to resample it myself first? (the only information I have in this regard is this email exchange years ago Re: [Freesurfer] anisotropic input data) If so which packages or technique is best?
Furthermore, between 0.7mm and 1mm, which is a safer bet? I know that downsampling to 1mm would lead to a loss in resolution, but would resampling to 0.7mm result in extra noise compared to 1mm? and is the extra resolution worth it?
Finally, I’ve also tried preprocessing my data with sMRIprep which has resulted in okay-ish results, which I’m assuming is due to the pre-freesurfer steps processing the data at the original resolution and freesurfer downsampling the image back to 1mm. I’ve also tried taking the sMRIprep only output and using the T1w image on fastsurfer, the results looked surprisingly reasonable but I’m worried that I might be missing something important.
Sorry if some of the questions seem basic and thank you for any responses.