Hi Kev, For the two scenarios you mentioned, here are my best estimates. I assume a healthy, adult population scanned with our recommended sequences. Any other type of dataset (pediatric, one with some type of pathology, one acquired long ago using a less than optimal sequence or coil, etc.) would likely differ by a lot in terms of time estimates.

1. fMRI on good quality 3T Maybe 25 out of 100 need some kind of intervention, including restarting a process that may have failed along the way. I would expect these 25 to have something terribly wrong with them to make them unacceptable for fMRI analysis but something that is most likely easily fixed by an automatic intervention. I'd estimate a half hour for each of the 25 or an hour for each to be extremely conservative.

2. Cortical thickness on good quality 3T Maybe 50-75 out of 100 need some kind of intervention, depending on which areas you might focus on if you're not only interested in global thickness. Edits on half of those will be very small, 15-30 minutes a case. Edits on the other half may be more time consuming, 30 minutes to 1 hour.

Allison

On Tue, 5 Mar 2013, Schumman Resonance wrote:

This is a question for the FS community in general, I gather there is variability around this, but rather than an exact number and time, I was hoping for an estimate on how many scans would I expect to require manual edits(any of the check points recommended in the tutorials) as well as an estimate of the time it takes to go through them.

For the sake of the argument, take a sample of 100 subjects.

Thanks,

Kev

_______________________________________________

Freesurfer mailing list

Freesurfer@nmr.mgh.harvard.edu

https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer