Back

Prefrontal neural dynamics for behavioral decisions and attentional control

Erez, Y.; Kadohisa, M.; Petrov, P.; Sigala, N.; Buckley, M. J.; Kusunoki, M.; Duncan, J.

2020-05-07 neuroscience
10.1101/2020.05.06.080325 bioRxiv
Show abstract

Complex neural dynamics in the prefrontal cortex contribute to context-dependent decisions and attentional competition. To analyze these dynamics, we apply demixed principal component analysis to activity of a primate prefrontal cell sample recorded in a cued target detection task. The results track dynamics of cue and object coding, feeding into movements along a target present-absent decision axis in a low-dimensional subspace of population activity. For a single stimulus, object and cue coding are seen mainly in the contralateral hemisphere. Later, a developing decision code in both hemispheres may reflect interhemispheric communication. With a target in one hemifield and a competing nontarget in the other, each hemisphere initially encodes the contralateral object, but finally, decision coding is dominated by the task-relevant target. These findings further suggest that exchange of information between hemispheres plays a key role when attentional competition resolves. Tracking complex neural events in a low-dimensional activity subspace illuminates integration of neural codes towards task-appropriate behavior, comprising a building block of learned task structure in the prefrontal cortex. AUTHOR SUMMARYFlexible adaptive processing of information is integral for everyday goal-directed behavior. To unravel dynamic representation of task-relevant information, we analyzed population activity of a primate prefrontal cell sample in a cued target detection task. In a low-dimensional neural subspace, with separate axes for cue, object identity and decision, trajectories showed initial coding of cue and object in the contralateral hemisphere, followed by coding of the behavioral decision across both hemispheres. With target and nontarget stimuli in opposite hemifields, the data suggest initial coding of the contralateral object in each hemisphere. Object coding is then rapidly shut off for the nontarget, and followed by bilateral coding of the target decision. The results bring detailed insight into task structure and information flow within and between the two frontal lobes as a decision is made and attentional competition is resolved.

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.