Nonlinear Spatial Integration Underlies the Diversity of Retinal Ganglion Cell Responses to Natural Images

How neurons encode natural stimuli is a fundamental question for sensory neuroscience. In the early visual system, standard encoding models assume that linearly filter incoming through their receptive fields, but artificial stimuli, such as contrast-reversing gratings, often reveal nonlinear spatial processing. We investigated to what extent processing relevant of images in retinal ganglion cells mice either sex. found linear field yielded good predictions responses flashed subset failed capture spiking activity many others. Cells with poor model performance displayed pronounced sensitivity fine contrast and local signal rectification dominant nonlinearity. By contrast, high-frequency classical test integration, was not predictor thus did variability integration under images. addition, we also observed class inverse tuning responding more strongly spatially homogeneous than structured stimuli. These findings highlight diversity nonlinearities crucial component understanding context SIGNIFICANCE STATEMENT Experiments have revealed types pool input signals nonlinearly. However, it still unclear how this when are Here analyze scenes large populations mouse cells. show influences some cells, were sensitive structure inside small group surprising Traditional analyses gratings predict