A computational model for characterizing visual information using both spikes and Local Field Potentials

How the brain maintains the stability of visual perception across saccade is a central question in systems neuroscience; accurately characterizing visual responses in the perisaccadic period is an important step towards understanding how the visual world is represented during saccades. Here, we develop a probabilistic model in the Generalized Linear Model framework to characterize and predict the pre-, trans-, and post-saccadic responses of single Middle Temporal (MT) neurons at the level of single-trial spike trains. We fit the MT spike response to a model, consisting of a set of stimulus kernels capturing neuron's spatiotemporal filtering properties, a set of saccade kernels capturing changes in stimulus sensitivity induced by eye movement, an offset kernel capturing time-varying baseline activity relative to saccade, a post-spike kernel capturing dependencies on spiking history, a set of coupling kernels capturing correlations between neurons, and finally an LFP kernel capturing effects of brain state in generating spike responses. The model, with parameters fit directly to data, accurately predicts responses to novel stimuli on individual trials during saccadic eye movements. The model can also be used to optimally decode time-varying visual information carried by MT responses, thus providing a tool for reading out visual information and understanding the representation of visual scene during eye movements.