Chaotic Signal Analysis of Parkinson's Disease STN Brain Signals

Techniques of signal processing from the field of non-linear time series analysis have been applied to microelectrode recordings from the subthalamic nucleus of patients suffering from Parkinson's disease. The recordings have been taken while the patients perform different language processing tasks from raw microelectrode recording data obtained from within the brain during surgery. Using classical chaotic signal analysis techniques, very complex behaviour is discerned but no robust trends for different tasks identified. However interesting behaviour is shown when the recordings are analysed using techniques from non-equilibrium statistical mechanics, indicative of different short term signal behaviour whilst performing the different language processing tasks. In particular, the evidence of discrimination between different linguistic tasks and left/right brain behaviour is revealed.