Fronto-parietal cortical circuits encode accumulated evidence with a diversity of timescales
Benjamin B. Scott*, Christine M. Constantinople*, Athena Akrami, Timothy D.Hanks, Carlos D. Brody and David W. Tank, Neuron, 2017
Decision making often involves evidence accumulation, in which newly acquired information is used to update existing beliefs and select future actions. Behavioral studies have suggested that the accumulation process is crucial to making good choices, whether in simple perceptual discriminations, or complex problem solving. How exactly does the brain combine past experience with new information to make decisions?
Two competing hypotheses have been proposed to describe how evidence accumulation might be implemented in neural circuits. The current leading model (homogeneous model) proposes that a specialized network of neurons all increase their rate activity the same time when evidence is presented. The more evidence that is presented, the greater the activity in these cells. In an alternative model (heterogeneous model), neurons do not become active all at once: some respond quickly and transiently to new evidence while others respond more slowly. In this model, evidence is collectively encoded by groups of neurons that respond in diverse ways to new information.
To distinguish between the two models of accumulation, Ben and Christine recorded from frontal and parietal cortex using two-photon calcium imaging techniques while rats performed a pulse-based accumulation of evidence task. In this task rats, viewed two LED that each produced an independent stream of randomly timed flashes. The rats were rewarded for selecting to the LED that had the greater number of flashes. In previous work (Scott, Constantinople 2015) Ben and Christine demonstrated that rats solve this task by an accumulation process, in which they estimated the number of flashes from each LED.
When Ben and Christine looked at the responses of frontal and parietal regions during the task they found that each pulse of evidence triggered a wave of activity that slowly spread through the cortex. These waves continued to propagate throughout the cortex until the choice was made. Although the response of each individual neuron was relatively brief, together they contained enough information to accurately predict the number of light pulses that the animal saw. The activity was best explained by the heterogeneous model of memory, and may reflect a strategy to construct memories that are stable over time but that can also be easily modified as new information becomes available.