Dynamic Analysis of Human Brain in the Pain State by Electroencephalography
Abstract Objective Pain is an unpleasant sensation that is important in all therapeutic conditions. So far, some studies have focused on pain assessment and cognition through different tests and methods. Considering the occurrence of pain causes, along with the activation of a long network in brain regions, recognizing the dynamical changes of the brain in pain states is helpful for pain detection using the electroencephalogram (EEG) signal. Therefore, the present study addressed the above-mentioned issue by applying EEG at the time of inducing phasic pain. Results Phasic pain was produced using coldness and then dynamical features via EEG were analyzed by the Recurrence Quantification Analysis (RQA) method, and finally, the Rough neural network classifier was utilized for achieving accuracy regarding detecting and categorizing pain and non-pain states, which was 95.25\(\pm\)4%. The simulation results confirmed that cerebral behaviors are detectable during pain. In addition, the high accuracy of the classifier for evaluating the dynamical features of the brain during pain occurrence is one of the most merits of the proposed method. Eventually, pain detection can improve medical methods.