October 5, 2012
Title: An Overview Of Some Recent Results In Psychiatric Patient Stratification Using Novel Data Analysis Methods
Speaker: Joseph Geraci, Toronto General Hospital and Queen’s University
Abstract: A strong desire exists to determine the efficacy of treatment for MDD (major depressive disorder) and other psychiatric conditions from first visit MRIs and clinical testing. I will present some preliminary work that is leading to this goal for rTMS treatments (repeated trans-cranial magnetic stimulation) for MDD. The same effort will be made to create an algorithm that scores the efficacy of pharmaceutical treatments. The methods include techniques from graph theory and machine learning. I will also present a novel machine learning algorithm called Butterfly that I have created over the last two years based on discrete dynamical systems. This talk will be technically elementary and no knowledge of any of the mathematical concepts will be assumed.