At LAMCIG, we develop computational methods to jointly analyze brain MR images arising from multiple modalities. We hope to fuse knowledge from cortical and subcortical anatomy, macroscopic brain connectivity, microscopic tissue properties and cellular dynamics to model neurological disease progression. As a parallel goal, we also explore effects of genetic control on imaging phenotypes and the subsequent possibility of genetic discovery, known more recently as “imaging genetics.” Our algorithm development relies heavily on tools from statistical learning and differential geometry.