Although we record and analyze different types of brain oscillations, our focus is on “gamma” oscillations (30-80 Hz). Our grand goal is to understand the neural mechanisms underlying high-level cognition. Gamma oscillations are modulated by high-level cognition, but can also be induced in the visual cortex by presenting certain visual stimuli such as bars and gratings, and gamma oscillations are highly dependent on the properties of the visual stimulus, such as its size, spatial frequency, contrast, etc. Further, these stimuli are linked with certain neural mechanisms, such as gain control, excitation-inhibition (E-I) balance, and normalization. Finally, a lot of modeling studies have shown how circuit-level implementation of certain mechanisms affects gamma oscillations. Therefore, it is possible to find “links” between gamma with (i) cognitive functions, (ii) mechanisms such as gain control and E-I balance, and (iii) their circuit-level implementations. We think that a thorough understanding of these links will allow us to understand the circuit mechanisms underlying high-level cognition.
Research Highlights
We studied how gamma power and center frequency change as the stimulus properties are modified, and from that try to understand the properties of the underlying network. For example, we found that presenting large (full screen) gratings induces two gamma oscillations, one between 40-70 Hz (“fast gamma”) that appears to be involved in local processing and a slower gamma between 25-40 Hz that is more global (Dinavahi, Shirhatti, Ravishankar and Ray, 2018, Journal of Neuroscience). Similarly, We found that gamma rhythm is highly color tuned, such that the presentation of long wavelength (red-ish) hues generates gamma rhythm of very large magnitude (Shirhatti and Ray, 2018, PNAS). These oscillations with severely diminished by the presentation of plaid stimuli that invoke a mechanism called “Normalization” (Das and Ray, 2021, Cerebral Cortex Communications) and in the presence of small stimulus discontinuities that change the E-I balance (Shirhatti, Ravishankar and Ray, 2022, PLoS Biology).
We have also modeled the properties of these oscillations such as their shape (Krishnakumaran, Raees and Ray, 2022, PLoS Computational Biology), burst characteristics (Krishnakumaran and Ray, Cerebral Cortex), and effect of discontinuities (Shirhatti, Ravishankar and Ray, 2022, PLoS Biology). Together, these studies show how E-I interactions are reflected in the properties of gamma oscillations.
