On the evolution of neural decisions from uncertain visual input to uncertain actions

Abstract

Behavior can be conceived as the result of a sequence in which the outcomes of perceptual decisions inform decisions on which action to take. However, the relationship between these processes, and spatiotemporal dynamics of the visual-to-motor transformation remains unclear. Here, we combined accumulation-to-threshold models and electro-magnetoencephalography, to trace neural correlates of sensorimotor decisions in space, time and frequency. We challenge the assumption of sequential decisions, with evidence that visuomotor processing unfolds through a continuous flow of information from sensory to motor regions. Action selection is initiated before regional visual decisions are completed. By linking behavior and physiology through theoretical decision models, we identify simultaneous forward and backward flow of information for visuomotor decisions between sensory and motor regions, in beta and gamma ranges. The model of integrated visuomotor decisions provides a powerful approach to investigate behavioral disorders that impair the ability to use sensory inputs to guide appropriate actions.