Back

The timescale and magnitude of aperiodic activity decreases with cortical depth in humans, macaques and mice

Halgren, M.; Kang, R.; Voytek, B.; Ulbert, I.; Fabo, D.; Eross, L. G.; Wittner, L.; Madsen, J.; Doyle, W. K.; Devinsky, O.; Halgren, E.; Harnett, M. T.; Cash, S. S.

2021-07-29 neuroscience
10.1101/2021.07.28.454235 bioRxiv
Show abstract

Cortical dynamics obey a 1/f power law, exhibiting an exponential decay of spectral power with increasing frequency. The slope and offset of this 1/f decay reflect the timescale and magnitude of aperiodic neural activity, respectively. These properties are tightly linked to cellular and circuit mechanisms (e.g. excitation:inhibition balance and firing rates) as well as cognitive processes (e.g. perception, memory, and state). However, the physiology underlying the 1/f power law in cortical dynamics is not well understood. Here, we compared laminar recordings from human, macaque and mouse cortex to evaluate how 1/f aperiodic dynamics vary across cortical layers and species. We report that 1/f slope is steepest in superficial layers and flattest in deep layers in each species. Additionally, the magnitude of this 1/f decay is greatest in superficial cortex and decreases with depth. We could account for both of these findings with a simple model in which superficial cortical transmembrane currents had longer time constants and greater densities than those in deeper layers. Together, our results provide novel insight into the organization of cortical dynamics, suggesting that the amplitude and time constant of local currents control circuit processing as a function of laminar depth. This may represent a general mechanism to facilitate appropriate integration of fast sensory inputs (infragranular) with slow feedback-type inputs (supragranular) across cortical areas and species.

Matching journals

The top 3 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.