Functional hyperalignment in pediatric fMRI
- Tamara Vanderwal, University of British Columbia
Brain scans from MRI machines are an important research tool for studying psychiatric and neurological disorders in children. MRI lets us study brain structure, and functional MRI (fMRI) lets us also study brain function. In fMRI, each child’s brain images need to be lined up in a special way so that proper comparisons and statistical tests can be done across everyone in the study. This important step, called alignment, happens in every fMRI study of every disorder. Usually the alignment is done using physical landmarks, meaning each brain is lined up according to structural features. In children, alignment is challenging because of differences in size and shape across brains. Recently, a new way of doing alignment has been developed. Instead of using structural landmarks, researchers have started using the functional signal to do the alignment. So far, this procedure—called “hyperalignment”—has worked better. For example, when tests are done using both methods, the hyperaligned data could predict age and other things about a person with better accuracy. This is sort of similar to using a new photo processing step that results in clearer photographs, so you can see details and recognize faces that were blurry before. Hyperalignment could have an enormous effect on pediatric fMRI studies. It has not yet been tested in children because it requires long brain scans with very low head motion for which children struggle to stay still. The current project overcomes this limitation by using movies in the scanner. The movies help children stay still, and also drive brain function in useful ways. These data will be innovative and unique, so will share them with other researchers. We will also use the data to test hyperalignment in children, with the goal of making it a method that could significantly improve pediatric fMRI research overall.