A Bayesian method to estimate single-trial event-related potentials with application to the study of the P300 variability

摘要: We propose a Bayesian method to extract single-trial event related potentials (ERPs). The is formulated in two stages. In the first stage, each of N raw sweeps processed by an individual "optimal" filter, where 2nd order priori statistical information on background EEG and unknown ERP is, respectively, estimated from pre-stimulus data obtained through multiple integration white noise process model which identifiable post-stimulus thanks smoothing criterion. Then, mean determined as weighted average filtered sweeps, weight inversely proportional expected value norm correspondent filter error. second single-sweep estimation dealt with within same framework, using previous stage response. successfully tested simulated then employed real aim investigating variability P300 component during cognitive visual task. A comparison other literature methods also performed. Results encourage further use proposed investigate if how diseases, e.g., cirrhosis, are associated differences variability.