Attention: Cortical Microcircuits
I studied Bioengineering at the University of Pennsylvania before coming to Stanford in 2007. I began a collaboration between the Brains In Silicon lab and Dr. Tirin Moore's neurophysiology lab, and have been pursuing experimental and modeling studies since then. My auxilliary research activities have included the Woods Hole "Methods in Computational Neuroscience" course (2009), the FENS-IBRO "Neural Coding in Sensory Systems" course (2012), many neuroscience courses at Stanford, and also assisting the teaching of two neuroscience courses here. For several years I organized Stanford's "Brain Day" program, bringing real brain samples to 7th grade classrooms around Palo Alto.
Selective attention is a key cognitive faculty that allows us to perceive and respond to only the chosen portion of the information that constantly and torrentially impinges on our brain via sensory systems. In the visual domain, this means that though our entire retinae are stimulated continually when our eyes are open, our higher visual centers typically ignore almost all of this information, instead responding to only those select parts that grab our attention. The improvements in our perception and memory of attended stimuli versus ignored stimuli have been examined extensively with psychophysical studies for many years, but the neural mechanisms underlying these perceptual improvements remain mysterious. Over the last twenty five years, neuroscientists have done much to elucidate the physiological correlates of these perceptual changes by recording the neural activity of the visual and frontal cortices, as well as of subcortical structures such as the superior colliculus and thalamus.
The present challenge, then, is to determine the mechanisms that generate these neural correlates of attention. Earning this understanding would improve our ability to understand what approaches are needed to effectively treat disorders of attention. We hope to make progress toward this goal with a combination of extracellular recordings in awake, behaving animals and computational models of spiking networks to explain the neurophysiological data and make novel predictions. In particular, we are carrying out recordings with linear array electrodes that allow for the simultaneous measurement of neural activity in multiple parts of the visual cortical microcircuit (i.e., in separate cortical layers and neuronal types). Our models will seek to include these layers and the pattern of interactions between them in order to better understand the microcircuit mechanisms underlying visual selective attention.
I am currently recording from an animal trained on a selective attention task, analyzing these data, and training another animal to perform the task. Simultaneously, I am developing spiking neuron models to account for the neural data on Neurogrid hardware, as demonstrated below and in the Neurogrid pages.