Entropy, complexity and chaos in brain rhythms

The classical approaches to analysis and interpretation of the brain rhythm, namely the EEG, are to employ non-parametric or parametric signal processing methods. These linear systems approaches to brain rhythm analysis have now given way to more advanced methodologies. These methods recognize that the brain rhythms are non-stationary and brain's responses to stimuli are non-linear. While spectral analysis has proved its value in sleep staging analysis, higher order spectral analysis has been useful in determining depth of anesthesia. Complexity analysis has been shown to discriminate neurological disorders such as schizophrenia. Chaotic dynamics have been observed in brain rhythms preceding or resulting from epileptic seizures. The concepts derived from information theory, including measures of entropy, have been useful in characterizing brain injury. Advanced signal processing has long been of interest in application areas such as diagnosis of brain disorders, epilepsy, sleep or anesthesia analysis, and more recently in brain-computer interfaces. An emerging application being developed by our group is monitoring brain's rhythm after neurological trauma or injury. Advanced quantitative analysis, based on the information and entropy analysis methods, has been used by our group to distinguish and characterize the injury response. This presentation will review the state of the art of brain rhythm analysis using the emerging signal processing methods and will especially help theoreticians targeting emergent, significant biomedical applications.