Auxiliary diagnosis of Schizophrenic Patients Based on Low-workload and Non-invasive Measurement

  • 謝 宗澔

Student thesis: Doctoral Thesis

Abstract

Schizophrenia (SZ) is a well-known and one of the least understood and costliest mental disorders in terms of human suffering and societal costs and it occurs in about 1% of the general world population The first-episode of this neuropsychiatric mostly occurs in 18-25 years old and has over 40% chance to become a lifetime illness that affects the patient's entire life According to the statistics of the Ministry of Health and Welfare of the R O C schizophrenia first time became the top ten cost of national health insurance in 2017 Each year government need pay about $12 7 billion NTD in patient’s diagnosis and treatment The main symptoms are hallucinations or delusions of perception and cognitive patient can identify the reality and hallucinations and then affect their outside characterization such as behavior language and emotion However the causes of schizophrenia are not yet clear current diagnosis is a time-consuming work which rely on the subjective report of patient-self or its relative person According to The Diagnostic and Statistical Manual of Mental Disorders (DSM) it may take more than 1-6 months to follow-up and confirm the diagnosis Paradoxically it is difficult to ask patient self-report their first-episode because they cannot identify the reality and hallucinations Usually this part relies on relative person observe their abnormal behavior language and emotion then reported It may become a lifetime illness The aim of this study is to propose the objectively distinguishing methods for identification Schizophrenia by analyzing physiological information It will help in reducing the time course of the diagnosis and provides additional information to the doctor The chosen measure of physiological information should be having two characteristics (1) easy to implement and not to cause discomfort to the subject; (2) In addition to auxiliary diagnosis the physiological information can also provide diagnosis or treatment information In this paper two non-invasive measurements of physiological information – Electroencephalography (EEG high-resolution in temporality) and functional magnetic resonance imaging (fMRI high-resolution in spatiality) were selected; For EEG measurement we refer the procedure of Liang et al (2013) to evoke the auditory-related potential (AEP) In this procedure the subject does not need to perform any task just close the eyes and passively listen to the sound stimulation of different complexity The patient's AEP amplitude will be significantly smaller than normal Further a feature selection strategy combines discrimination and correlation analysis is also proposed to select key features and remove redundancy Two AEP components amplitude of N1 evoked by chord stimuli and amplitude of P2 evoked by interval stimuli from the frontal lobe were screened and fed to the linear discriminate analysis (LDA) for classification The accuracy reaches 83 33% through leave-one-out cross-validation from 12 SZ and 12 healthy subjects; on the other hand fMRI is also a convenient non-invasive measurement However the environmental noise let AEP procedure cannot function Hence we analyzed the resting state fMRI of 72 normal people and 69 patients with mental disorders from the public database - COBORE Durning rs-fMRI recording participants are typically asked to rest quietly with their eyes open or closed for 5-10 minutes and without performing any tasks The whole brain network complexity and regional homogeneity of patients were significant lower than normal people Accuracy of the proposed linear-SVM can reach 73 05% through leave-one-out cross validation Both our proposed measurements provide over 70% correction and the recording methods are simple and easy to implement More important the measurements do not bring workload or strong discomfort to subject Moreover both measurements were point out the frontal lobe and temporal lobe were the key role of dysfunctional in schizophrenia It is expected to be a useful tool to help us understand abnormalities of brain function and a potential biomarker to plane the treatment strategy in schizophrenia Moreover it has high potential in simplifying the AEP procedure and develop portable diagnostic devices for the specific region
Date of Award2018 Nov 5
Original languageEnglish
SupervisorSheng-Fu Liang (Supervisor)

Cite this

'