The Non-linear Dynamic Analysis Of EEG Signals In Schizophrenia Patients During Manipulation Of Cool Executive Function Tasks

Schizophrenia is a more serious mental illness characterized by unknown etiology and complex symptoms,which is difficult to be diagnosed.The diagnosis of schizophrenia largely depends on the psychiatrists’ clinical experience as well as manuals of classification of diseases at home and abroad(CCMD-3,DSM-IV-TR and ICD-10 et al),but without one definite objective standard.With the development of electrophysiology and digital signal processing technology,the analysis of EEG signals applied to the diagnosis of schizophrenia has attracted many researchers.Some studies believe that due to the characteristics of nonlinearity,complexity and dynamic,EEG signals produced in the cerebral cortex are not fully analyzed by traditional linear analysis methods.In this study,the nonlinear dynamic analysis is utilized to effectively character EEG signals in schizophrenia patients,which is expected to establish a relatively objective standard in the diagnosis of schizophrenia.The pathogenesis of schizophrenia is still unclear,but studies suggest that schizophrenia has a variety of cognitive disorders,the main of which is executive dysfunction.Executive functions have been divided into two types: cool executive function,an abstract cognition,which reflect function in the dorsolateral prefrontal cortex(DL-PFC)and may be associated with relatively abstract and decontextualized tasks;hot executive function,with a high degree of emotional involvement.Cool executive function unrelated to emotional arousal is more objective and suited to study executive function in schizophrenia patients,because the emotional reaction of schizophrenia patients is often inconsistent with inner experience.Therefore,this study chose cool executive tasks as stimulus tasks.This study investigates cool executive function using fractal dimension,C0 complexity,and approximate entropy of EEG signals in seventeen first-episode schizophrenia patients and seventeen healthy controls matched for sex,age,and education background duringmanipulation of three cool executive function tasks with varying degrees of complexity:Trail Making Test-A(TMT-A),Trail Making Test-B(TMT-B),and the Tower of Hanoi task,and analyzes the difference of nonlinear features of EEG signals between patients and controls.Moreover,statistical analysis and support vector machine classifier are applied to the nonlinear features,and the results show that EEG signals in patients are more irregular and more complex,especially during manipulation of TMT-B,approximate entropy and fractal dimension can explain the difference of nonlinear features from perspectives of the irregular extent of time series and structural complexity,and good classification results,but the characterization performance of C0 complexity is not ideal.From the results of this study,the nonlinear dynamic analysis of EEG signal in schizophrenia patients combined with cool executive tasks may be regarded as a research approach to objectively reveal cognitive dysfunction of schizophrenia patients so that it can provide new reference for objective diagnosis and treatment of schizophrenia patients in the future.