Application of Generalized Linear Model in Neural Electrophysiological Signal Analysis
During advanced functional activities in the brain,neurons show distinct firing features in different brain regions,featuring different physiological states.For instance,neurons in the hippocampus,which are involved in cognitive activities such as learning and memory,show typical sharp-wave ripple(SWR)associated activities.Using multichannel electrophysiological recording technique,the extracellular electrical signals were recorded and analyzed in this study.Three physiological states(including awake,rapid eye movement sleep and non-rapid eye movement sleep stages)were classified by rhythmic local field potentials(LFPs)in the hippocampus.Meanwhile,a method based on the hippocampal SWRs was introduced to study the firing features of neuron in the hippocampus and cortex.Finally,a method for predicting neural signals by using a 5-fold cross-validated generalized linear model was presented,in which the matrices of the neuronal population firing within a specific time window were considered as the independent variables;the vectors of neuronal spike number or the SWR occurrence within that or other time windows were the dependent variables.These methods could predict the numbers of neuronal firing or the occurrence of SWR events.This study will be helpful to analyze and decode the neural signal patterns of brain regions under diverse physiological states,especially to provide a methodological basis for studying learning and memory processes and the related interactions between brain regions.
neural signalgeneralized linear modelsharp-wave ripplehippocampus
万方数据
维普
