Video From Applied Physics/Physics Colloquium: Mark Schnitzer - "Large-scale optical imaging studies of neural coding and computation in the neocortex"
Department of Physics
APPLIED PHYSICS/PHYSICS COLLOQUIUM
Tuesday, May 24, 2022 3:30 p.m. on campus in Hewlett Teaching Center, Rm. 200
Refreshments served in Varian lobby at 4:45 p.m.
Face coverings required in classrooms
Zoom link: https://stanford.zoom.us/j/96200437797
Password: email email@example.com for password.
"Large-scale optical imaging studies of neural coding and computation in the neocortex"
Optical techniques have become central to research at the forefront of brain science and are still rapidly increasing in their breadth and importance to the field. For instance, the United States Brain Research through Advancing Innovative Neurotechnologies (BRAIN) Initiative has as one of its priorities the aim of fostering continued innovation in this domain. I will present recent advances in optical brain imaging, which have allowed the visualization of large-scale neural coding in behaving animals as well as optical readouts of neuronal voltage oscillations. I will then describe how optical brain imaging has provided key insights into a fundamental problem first highlighted by John von Neumann, that of explaining how the brain achieves high-fidelity coding and computation despite the highly stochastic dynamics of individual neurons. Imaging studies at Stanford of the mouse visual cortex have revealed that dynamical fluctuations are correlated across large populations of individual neurons and across time-scales from seconds to days, enabling signal decoding schemes that use the statistical structure of this correlated variability to improve the reliability of signal transmission. Moreover, in mice performing a visual discrimination task, the neocortex encodes visual data and the mouse's task responses in orthogonal, non-interfering communication streams. Overall, by providing experimental access to the dynamics of thousands of neurons in behaving animals, optical techniques are uncovering the brain's basic computational principles.