Classification of single-trial neural responses to cutaneous stimulation

Given the response of a population of neurons to a set of stimuli (or events), classification methods have been used to determine which stimulus generated the response on a single-trial basis. Various linear and non-linear methods have been used to solve this classification problem with surprising little performance differences. We therefore developed a PSTH-based classification method with the aim of maximizing computational efficiency rather than classification performance. This method was tested by applying it to populations of neurons simultaneously recorded from the somatosensory cortex of the rat during cutaneous stimulation of the forepaw and the results were compared to linear discriminant analysis (LDA), which has been considered the "standard". This method performs at least as well as LDA and it is particularly efficient for complete cross-validation designs.