Abstract:

Despite a rich literature on the consequences of selective attention to behavior and to neural responses, as well as a wealth of models of competitive selection, little is known about the circuit mechanisms by which the brain actually selects the highest priority stimulus as the next target of spatial attention. Following a first principles approach that breaks down selection into computational primitives, we recently discovered neural circuit mechanisms for categorical (robust-to-noise) selection, location-invariant selection (selection at all possible pairs of locations), and flexible selection in the avian midbrain. I shall describe some of these findings, and in doing so, highlight (a) the identification of a new form of population coding in the brain, namely, combinatorially optimized coding (relating to classic np-hard optimization problems), as well as of (b) a novel, multi-holed donut-like organization of inhibition in the brain. I shall then switch gears and describe our development of primate-like behavioral paradigms for the study of endogenous as well as exogenous control of visuospatial selective attention in freely behaving mice. I shall end with results from our ongoing efforts involving opto-/chemo-genetic manipulations and endoscopic calcium imaging in mice that are aimed at identifying circuit mechanisms of selection for spatial attention in mammals. 

Bio:

Shreesh Mysore has a Bachelor’s degree in Mechanical Engg (IIT Madras), Master’s degrees in Industrial Engg and Mathematics (Penn State), and a PhD in Control and Dynamical Systems (Caltech). He obtained postdoctoral training in Neurobiology (at Stanford) before moving to Hopkins to start his research group. A common thread through this diverse background is a long-standing research interest in intelligent systems – first in artificial intelligence and robotics (from his undergraduate days through the first half of his PhD), and then in biological intelligence (the rest of his PhD to date). His lab studies neural circuit, computational and coding principles underlying complex behaviors and cognitive function including attention, decision-making, sensorimotor processing, and collaboratively, affective function. He is particularly intrigued by a comparative perspective – how do brains of different species solve similar behavioral challenges? A major goal of his lab’s research is to help develop novel, targeted therapeutics for psychiatric disorders of attention and executive function. A (secret) wish is to return to robotics, armed with insights from experimental neuroscience, and to help build novel and powerful AI. His work is funded by the NIH (NEI, NICHD, NINDS).

Abstract:

 Free boundary problems are ubiquitous in science and engineering but their numerical solutions are challenged by the fact that (1) smaller scales influence larger one in a non-trivial manner, which demands efficient adaptive gridding; (2) boundary conditions, including sharp jump conditions, must be imposed on a moving and irregular boundary and (3) nonlinearity requires advanced numerical schemes. I will present a framework that seeks to address these challenges and I will present some of their applications in the field of fluids and materials.

Biography:

Professor Gibou is a faculty member in the Department of Mechanical Engineering, in the Department of Computer Science and in the Department of Mathematics at the University of California, Santa Barbara. He is also a core faculty member in the Computational Science and Engineering program. He received his PhD from the Applied Mathematics Department at UCLA, and did his post-doctoral research in the Departments of Mathematics and Computer Science at Stanford University. He was awarded an Alfred P. Sloan Fellowship in Mathematics, the Regent’s Junior Faculty Fellowship, an NSF Mathematical Sciences Postdoctoral Fellowship and the Robert Sorgenfrey Distinguished Teaching award. Professor Gibou is on the Editorial Board of the Journal of Computational Physics .

Over the last decades, the fabrication of three-dimensional (3D) tissues has become commonplace. However, conventional 3D fabrication techniques are limited in their capacity to produce complex tissue constructs with the required precision and controllability that are needed to replicate their in vivo counterparts. To this end, 3D bioprinting offers great versatility in the fabrication of biomimetic volumetric tissues that are structurally and functionally relevant. The technology enables precise control of the composition, spatial distribution, and architecture of bioprinted constructs facilitating the recapitulation of the delicate shapes and structures of target organs and tissues. This talk will discuss our recent efforts in developing a series of advanced 3D bioprinting strategies along with various cytocompatible bioink formulations. These platform technologies, when combined with microfluidic systems, are likely to provide new opportunities in constructing functional tissues to facilitate regeneration as well as in generating microtissue models for promoting personalized medicine.