I was invited to give a presentation at the Oregon Health and Science University Neurology Grand Rounds in October 2009, where I focused on explaining how statistical models and clinical models intersect, interact, and support one another. 

This talk modeling4stat_clin.inf8oct09.ppt  was a shortened version of an invited lecture at Harvard Medical School in May 2009, which (actually!) led to one professor standing up at the end and declaring, "I am convinced!" 

StatMethodsforVarSel&modeling_brainaging.boston17jun09.ppt