Publications

Question:We previously showed that spevtral differentiation can identify specific brain areas that might be more involved in subjective perception. How does this metric, spectral differentation, depend on the timescale of observation, or change from one brain region to another?

Impact: We used the Allen Brain Observatory Visul Coding Neuropixels dataset to analyze the spectral differentiation of activity of ~50,000 neurons. We found that differentiation of single neurons is maximized at a timescale of ~100 ms, similar to the timescales of conscious perception; and the differentiation of activity of the entire visual cortex most faithfully reflects ‘conscious perception’ compared to individual areas in the visual cortex.

SR Gandhi, WGP Mayner, W Marshall, YN Billeh, C Bennett, SD Gale, C Mochizuki, JH Siegle, S Olsen, G Tononi, C Koch, A Arkhipov

Frontiers in Computational Neuroscience (2023)

Question:The claustrum, a centrally connected brain region, has been hypothetized to be involved in consciousness. How does the cortex respond to optogenetic stimulation of the mouse claustrum?

Impact: We found that unlike a few similar recent studies that suggest claustrum stimulation uniformly inhibits the cortex, claustrum stimulation can actually lead to a wide variety of responses in the cortex, depending on the cortical area and cell types observed. We characterized these responses in depth in this manuscript.

EG McBride, SR Gandhi, JR Kuyat, DR Ollerenshaw, A Arkhipov, C Koch, SR Olsen

Neuron (2022)

Question:Decoding is an extrinsic approach - the experimenter knows what is shown to the subject, and then asks- activity in what brain region can predict what was shown? But what is shown can be very different from what was consciously perceived by the subject. Can a new metric (spectral differentiation) predict what was consciously seen by the subject?

Impact: While decoding neural responses from any of the observed visual areas in mice can predict what was shown to the mouse, spectral differentiation in only specific visual areas (AL, AM) and layers (L2/3) is increased for ethologically ‘meaningful’ compared to ‘meaningless’ stimuli. Could it possibly be predicting what the mouse actually saw, not just what it was shown?

WGP Mayner, W Marshall, YN Billeh, SR Gandhi, S Caldejon, A Cho, F Griffin, N Hancock, S Lambert, EK Lee, JA Luviano, K Mace, C Nayan, TV Nguyen, K North, S Seid, A Williford, C Cirelli, PA Groblewski, J Lecoq, G Tononi, C Koch, A Arkhipov

ENEURO.0280-21.2021 9 (1) (2022)