Análise de dados intracranianos em humanos: utilizando metodologia simbólica para caracterizar diferentes processos cognitivos

Abstract

Understanding how the brain processes information, perceives the world, and creates models to act upon it is one of the main goals of neuroscience. To address this question, we analyze intracranial electrocorticography (ECoG) data from five human subjects performing a visual search task, using information theory quantifiers. The Bandt-Pompe symbolization methodology is employed to associate a probability distribution function with each time series, allowing the computation of entropy and complexity at different time scales. Our results show that different types of visual tasks can be distinguished in the multiscale entropy-complexity plane, suggesting that information theory quantifiers can serve as a valuable tool to differentiate brain signals in specific cortical regions. Moreover, by applying a multiscale approach and incorporating time delays to modify the data sampling rate, we estimate the most relevant time scales for processing information associated with each cognitive task.