Back

Cortex-wide characterization of decision-making neural dynamics during spatial navigation

Haley, S. P.; Surinach, D. A.; Nietz, A. K.; Carter, R. E.; Zecker, L. S.; Popa, L. S.; Kodandaramaiah, S. B.; Ebner, T. J.

2024-10-23 neuroscience
10.1101/2024.10.23.619896 bioRxiv
Show abstract

Decision-making during freely moving behaviors involves complex interactions among many cortical and subcortical regions. However, the spatiotemporal coordination across regions to generate a decision is less understood. Using a head-mounted widefield microscope, cortex-wide calcium dynamics were recorded in mice expressing GCaMP7f as they navigated an 8-maze using two paradigms. The first was an alternating pattern that required short term memory of the previous trial to make the correct decision and the second after a rule change to a fixed path in which rewards were delivered only on the left side. Identification of cortex-wide activation states revealed differences between the two paradigms. There was a higher probability for a visual/retrosplenial cortical state during the alternating paradigm and higher probability of a secondary motor and posterior parietal state during left-only. Three state sequences (motifs) illustrated both anterior and posterior activity propagations across the cortex. The anterior propagating motifs had the highest probability around the decision and posterior propagating motifs peaked following the decision. The latter, likely reflecting internal feedback to influence future actions, were more common in the left-only paradigm. Therefore, the probabilities and sequences of cortical states differ when working memory is required versus a fixed trajectory reward paradigm.

Published in Nature Communications (predicted rank #1) · training set

Matching journals

The top 4 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.