Hi Susana,

This is a common issue, I am sure if you try a threshold of (lets say) 0.2, you will avoid the skull. Note, generally BET is used for registration purposes, and therefore, you do not really care if your missing some brain. Perhaps others have more robust methods, but generally I have always been able to correct this issue by changing the -f option. Also, I find using the -R option helps a ton.

Bryson

On Tue, Mar 24, 2015 at 9:42 AM, Susana Santos Maia <[log in to unmask]> wrote:

Hi everyone,
I am fairly new to fMRI data analyses. I am having some problems regarding performing BET extraction in the structural images. I have played around with the several options BET offers, and also by changing the threshold level from the default 0.5, but I can not get a good extraction for one participant. I get an extracted image with a lot of non-brain (please see file attached). I would be very thankful if someone could give any advice on how to sort this.

Best regards,

    Susana Maia
    PhD student, Psychology department
    Room A105, Heslington, York, YO10 5DD, UK
    Email: [log in to unmask]
    Tel: +44 7456555876

    EMAIL DISCLAIMER http://www.york.ac.uk/docs/disclaimer/email.htm

Please consider the environment before printing this e-mail.