FHSS-SO-RCSV-UBSN Joint Distinguished Lecture - How Are Objects and Scenes Represented in the Brain? Applications of Differential Geometry to Human Retinotopic Maps

One fundamental question in philosophy is: How do we perceive the world? From physiology, we know that objects and scenes are first projected to the retina and then represented in multiple visual cortical areas. The goal of our research is to specify the exact geometric transformations of images on the retina (“the world”) in the human brain. Animal studies have found that although they do not preserve the exact shape of visual input on the retina, cortical representations preserve topology - they can be unfolded, stretched, compressed, shifted, and rotated (without cutting/tearing) such that they are aligned to the exact shape of the visual input. For humans, recent advances in high field functional magnetic resonance imaging (fMRI) have made it possible to generate in vivo cortical representations of retina inputs with great details. We developed a new framework based on advanced differential geometry to rigorously characterize the transformations of retinal images in the brain. The new framework will help us address some fundamental philosophical questions and can be applied to evaluate cortical transformations for individuals with and without eye diseases as well as shed new light on the functional organization of the visual system.