Contribution of sensory encoding to measured bias

Abstract

Perceptual decision-making is a complex process that involves sensory integration followed by application of a cognitive threshold. Signal detection theory (SDT) provides a mathematical framework for attributing the underlying neurobiological processes to these distinct phases of perceptual decision-making. In particular, SDT reveals the sensitivity (d) of the neuronal response distributions and the bias (c) of the decision criterion, which are commonly thought to reflect sensory and cognitive processes, respectively. However, neuronal representations of bias have been observed in sensory areas, suggesting that some changes in bias are due to effects on sensory encoding. To directly test whether sensory encoding can influence bias, we optogenetically manipulated neuronal excitability in primary visual cortex (V1) during a detection task. Increasing excitability in V1 significantly decreased behavioral bias, while decreasing excitability had the opposite effect. To determine whether this change in bias is consistent with the effects on sensory encoding, we made extracellular recordings from V1 neurons in passively viewing mice. Indeed, we found that optogenetic manipulation of excitability shifted the neuronal bias in the same direction as the behavioral bias, despite using a fixed artificial decision criterion to predict hit and false alarm rates from the neuronal firing rates. To test the generality these effects, we also manipulated the quality of V1 encoding by changing stimulus contrast or inter-stimulus interval. These stimulus manipulations also resulted in consistent changes in bias measured both behaviorally and neuronally. Thus, changes in sensory encoding are sufficient to drive changes in bias measured using SDT.