Understanding Form and Function in the Human Brain through Diffeomorphic Image Normalization

We estimate biologically plausible, high-dimensional mappings between individuals in order to uncover complex patterns and relationships in anatomy and physiology.  The majority of our work is on the brain.  However, our methods are often applied to heart and lung. 

This group strives to improve image normalization technology through advanced mathematics, prior knowledge and the potential to include user input. We apply these methods to analyzing interesting medical conditions, answering questions of epidemiological interest, improving image-based data mining of genetic expression and also imaging studies that relate to questions about human evolution. As often as possible, we hope to translate our techniques to the scientific community.

One of our main specialties is using the diffeomorphic space in image registration. We also use the finite element method for flattening and normalization, B-Splines and medial models. Diffusion tensor imaging is also a major focus, as is the fusion and use of multiple modalities in building a complete description of subjects of interest.