In most neuro imaging research it is needed to study specific areas of the brain instead of the wholevolume. For me, as many other researchers, it is very difficult or impossible to locate by hand some specific brain areas. And it is not desired to do conclusions over an area that is not correctly located, because this conclusions will end up corresponding to some other part of the brain. So, is very important to use a good tool for registration. The following papers explain why ANTs is one of the best tools for Brain Registration:
http://www.mindboggle.info/ papers/evaluation_ NeuroImage2010/Evaluation_ Klein_NeuroImage2010.pdf
http://www.mindboggle.info/ papers/evaluation_ NeuroImage2009/Evaluation_ Klein_NeuroImage2009.pdf
There are no “best parameters”for every registration, it mostly depends on the resolution of the images and the moving and fix spaces as well as many other characteristics. However, I think that there should be some basic recommendations for running a correct corregister, because it would be a waste of time to try on with hundred possibilities for each parameter. As a new ANTs user I struggled a lot trying to find out which could be the best options to be used, because every paper I read would say that the best parameters were different depending each specific situation. But still, for a new user is difficult to guess from zero the best options. So finally, after reading ANTs manual and with personal communication with Dr. Klein, I did a resume that I hope can help as a basis for other people that are learning about ANTs as well.
The file ants.h that can be found on my web page contains my recommended parameters for doing brain registration from an atlas to a subject's space. It can be executed directly from the command line if you have ANTs installed in your computer. You can find there as well an explanation on how the parameters were chosen.
Visit www.monicagiraldochica.com (go to 'My PhD Project' > Project Progress > Files)
http://www.mindboggle.info/
http://www.mindboggle.info/
There are no “best parameters”for every registration, it mostly depends on the resolution of the images and the moving and fix spaces as well as many other characteristics. However, I think that there should be some basic recommendations for running a correct corregister, because it would be a waste of time to try on with hundred possibilities for each parameter. As a new ANTs user I struggled a lot trying to find out which could be the best options to be used, because every paper I read would say that the best parameters were different depending each specific situation. But still, for a new user is difficult to guess from zero the best options. So finally, after reading ANTs manual and with personal communication with Dr. Klein, I did a resume that I hope can help as a basis for other people that are learning about ANTs as well.
The file ants.h that can be found on my web page contains my recommended parameters for doing brain registration from an atlas to a subject's space. It can be executed directly from the command line if you have ANTs installed in your computer. You can find there as well an explanation on how the parameters were chosen.
Visit www.monicagiraldochica.com (go to 'My PhD Project' > Project Progress > Files)