Coding perceptual decisions: emergent signaling properties in cortical circuits
Spiking activity in single neurons of the primate visual cortex has been tightly linked to perceptual decisions. Any mechanism that reads out these perceptual signals to support behavior must respect the underlying neuroanatomy that shapes the functional properties of sensory neurons. Spatial distribution and timing of inputs to the next processing levels are critical, as conjoint activity of precursor neurons increases the spiking rate of down-stream neurons and ultimately drives behavior. In a recent reveiw I set out how correlated activity might coalesce into a micropool of task-sensitive neurons signaling a particular percept to determine perceptual decision signals locally and for flexible interarea transmission depending on the task context (Krug 2020 Annual Reviews of Vision Science). As data from more and more neurons and their complex interactions are analyzed, the space of computational mechanisms must be constrained based on what is plausible within neurobiological limits.
In area V5/MT, we recorded interneuronal correlations at local sites between a single neuron and the surrounding multi-unit activity, which likely constitutes a functional signature of the surrounding column. This interneuronal correlation measures scales with the strength of the perceptual choice signal of the single unit recorded at this cortical site (Wasmuht et al. 2019 Scientific Reports). We demonstrated this for perceptual decisions about 3D stereo-motion figures, but theoretical accounts of motion discrimination have also proposed that increased interneuronal correlations are associated with significant perceptual decision signals in visual cortex (Haefner et al. 2013, Shadlen et al. 1996). Over the course of a trial, as a perceptual decision about 3D stereo–motion signals is formed, the strength of local interneuronal correlations between spatially (and functionally) close neurons rises. The neuronal dynamic of increasing correlation appears to support larger perceptual choice signals, measured as CP (Wasmuht et al. 2019, Wimmer et al. 2015).