Objective quantification of selective attention in schizophrenia a hybrid TMS – EEG approach

W Azlan, Wan Amirah (2017) Objective quantification of selective attention in schizophrenia a hybrid TMS – EEG approach. Masters thesis, Universiti Teknikal Malaysia Melaka.

	Text (24 Pages) Objective Quantification Of Selective Attention In Schizophrenia ; A Hybrid TMS-EEG Approach.pdf - Submitted Version Download (230kB)
	Text (Full Text) Objective quantification of selective attention in schizophrenia a hybrid TMS – EEG approach.pdf - Submitted Version Restricted to Registered users only Download (1MB)

Abstract

Schizophrenia is a brain disorder that exhibits effects on perception, way of thinking and behavior. Often, schizophrenia patients suffer from attention deficiency. Currently clinical interview is used to diagnose schizophrenia by doctors. There is no alternative way to diagnose schizophrenia in present. Thus, an objective approach by employing transcranial magnetic stimulation combined with electroencephalogram (TMS-EEG) is proposed. The aim of the study is to quantify objectively the neural correlate of selective attention that reflected in auditory late responses (ALRs) using signal processing techniques. TMS provides a means of stimulating neuronal structures within the cortex using brief time-varying magnetic pulses generated by a coil positioned over the scalp. Integrating it with electroencephalogram provides real-time information on cortical reactivity and connectivity through the analysis of TMS evoked potentials or induced oscillations. In this project, auditory oddball paradigm was used throughout the experiment. The experiment involved three sessions; 1) without TMS, 2) single pulse TMS (sTMS) and 3) repetitive TMS (rTMS). All sessions were conducted in attended (attention) and unattended (no attention) conditions. It is found that the amplitude of the grand averaged of ALR (the N1-P2 wave) is higher in control compared to schizophrenia in without TMS session at both conditions. However, the amplitude of ALR in schizophrenia subjects is higher than control subjects in sTMS and rTMS at both conditions. The attention level measure, i.e., the Wavelet Phase Stability (WPS) was used to extract and quantify the neural correlates of auditory selective attention reflected in ALRs. In particular, Complex Morlet was implemented (scales 50-100 corresponding to 4-8Hz). There are significant differences of the ALR between schizophrenia and control groups in without TMS (p<0.05) and sTMS at the attended condition (frontal electrodes). Meanwhile at the unattended condition, Significance difference is found between two groups of the subjects in without TMS but no significant difference in sTMS (frontal electrodes). Particularly, the WPS of controls are larger than schizophrenia patients for without TMS and sTMS at attended for frontal electrodes. These results were consistent for temporal electrodes. It is worth to note that the phase stability of ALR in single pulse TMS is lower than without TMS for controls during attended but showed reversed pattern in unattended. Besides, it is found that a large phase stability difference between without TMS and sTMS in schizophrenia (frontal and temporal electrodes) at unattended compared to attended. For control subjects, this difference is small at frontal and temporal electrodes in both conditions. In a further investigation, the C4.5 decision tree algorithm was implemented to classify the N1-P2 wave of control and schizophrenia subjects elicited by sTMS and rTMS. Four features (energy, power, variance and entropy) were extracted by continuous wavelet transform (CWT). The result shows high classification accuracy which is above 83% in all three sessions at both attended and unattended conditions. In conclusion, the combined TMS-EEG approach shows a promising way to study the selective attention in schizophrenia. By successfully quantifying the neural correlates of auditory selective attention reflected in ALRs using the WPS and discriminating the control and patient groups using C4.5 decision tree provides an objective way to diagnose schizophrenia in compliment to the current subjective method.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Signal processing, Electroencephalography, Data processing, Schizophrenia, Hybrid TMS-EEG
Subjects:	Q Science > Q Science (General) Q Science > QP Physiology
Divisions:	Faculty of Electronics and Computer Engineering
Depositing User:	Mohd Hannif Jamaludin
Date Deposited:	15 Mar 2018 06:26
Last Modified:	13 May 2022 11:26
URI:	http://eprints.utem.edu.my/id/eprint/20539
Statistic Details:	View Download Statistic

Actions (login required)

View Item