Hello,
You need to determine the threshold visually, and it is fine if you use different values for different images. The values themselves are not important, it is only important to get a good brain extraction. Since there are lots of variations and parameter settings for MRI acquisitions, it is not possible to predict in advance the best threshold. That is why it needs to be selected based on visual inspection.
All the best,
Mark
P.S. Please don't send multiple emails to the list containing the same question as it makes it hard to keep track of them.
On 22 May 2014, at 14:51, SUBSCRIBE FSL ShiXinfa <[log in to unmask]> wrote:
> Dear FSL experts:
>
>
> When extracting brain in DTI, I find that the different threshold have different extraction results,0.2 is better among them.
> Therefore I have the following questions:
>
> 1. When extracting brain , do I have to choose a specified threshold first and then visually check the extraction result ?
> If so , how much is usually the specified threshold ?
>
> 2. Do you have any better suggestion to tell me that how to choose threshold ?
>
> Thank you very much .
|