Biological Subtyping in Psychiatry: Hope, Controversy, and A New Method

This lecture explores the recent advancements in using machine learning for biological subtyping in psychiatry, focusing on major depressive disorder and autism spectrum disorder. The instructor discusses the promise and controversy surrounding these methods, emphasizing the potential for personalized diagnosis and treatment. Various neuroimaging approaches and clustering methods are presented, highlighting the identification of distinct patient subgroups. The talk also introduces a new method that significantly enhances existing approaches for patient subgroup identification using multiomics and neuroimaging data.