ntustison
commented
1 year ago Template creation is constrained by the requirement of having spatial correspondences across the population used to construct the template. We had a related issue in our cortical thickness paper where we were using multiple datasets each with their own defacing protocol. We also needed the non-brain tissue for normalization purposes so we ended up creating multiple templates
. In your case, you might have to do something similar, especially if you need it for normalization purposes.
I need to create a template from the population using multimodal approaches something like this article. As I am using open source datasets like open-euro , all of them are defaced. therefore the template creation would be optimal on the brain tissues if I remove the skull, however I am also interested in optimizing the template for none brain tissues as I want to use those skull-muscles for normalization proposes! How do you solve this problem? Any help really appreciated
Best Rouhi