A Model for the Peak-Interval Task Based on Neural Oscillation-Delimited States

Abstract

Specific mechanisms underlying how the brain keeps track of time are largely unknown. Several existing computational models of timing reproduce behavioral results obtained with experimental psychophysical tasks, but only a few tackle the underlying biological mechanisms, such as the synchronized neural activity that occurs throughout brain areas. In this paper, we introduce a model for the peak-interval task based on neuronal network properties. We consider that Local Field Potential (LFP) oscillation cycles specify a sequence of states, represented as neuronal ensembles. Repeated presentation of time intervals during training reinforces the connections of specific ensembles to downstream networks. Later, during the peak-interval procedure, these downstream networks are reactivated by previously experienced neuronal ensembles, triggering actions at the learned time intervals. The model reproduces experimental response patterns from individual rats in the peak-interval procedure, satisfying relevant properties such as the Weber law. Finally, we provide a biological interpretation of the parameters of the model.

Interval Timing: Modeling the break-run-break pattern using start/stop threshold-less drift-diffusion model

“…Most interval timing models focused solely on reproducing the average response curves without looking at individual trial statistics such as the start and stop time and their correlations. The only exception we found are SET, TDDM, and the Neural Oscillation States model [Church et al, 1994, Luzardo et al, 2017b, Varella et al, 2019.…”

“…Instead, they model average population curves. The only exception we found is [Varella et al, 2019]. In this model, a network goes through a sequence of states.…”

“…Packet theory [Kirkpatrick, 2002, Kirkpatrick andChurch, 2003] is based on bouts of response that increase in likelihood as the likelihood of reward increases. Other approaches are based on neural oscillation, such as Striatal Beat Frequency (SBF) [Matell and Meck, 2000, Buhusi and Meck, 2005, Oprisan and Buhusi, 2011, Oprisan et al, 2014 and a more recent model based on Neural Oscillation States [Varella et al, 2019]. Other adaptive models include recurrent neural network approaches [Hardy andBuonomano, 2018, Raphan et al, 2019] and behavioral Learning to Time Theory (LeT) [Machado, 1997, Pinheiro de Carvalho et al Finally, in [Hasegawa andSakata, 2015], they developed a multi-clock model based on LeT and SET, that provided similar results to SET.…”