Assessment of internet gaming disorder severity based on autonomic nervous system using convolution neural network model

Title
Assessment of internet gaming disorder severity based on autonomic nervous system using convolution neural network model
Author
홍성준
Alternative Author(s)
SungJun Hong
Advisor(s)
김인영
Issue Date
2018-08
Publisher
한양대학교
Degree
Doctor
Abstract
Internet gaming disorder (IGD) is a new concept in the field of psychiatry, which means socially negative consequences due to excessive use of computer games. The Internet gaming disorder, like the general impulse control disorder, is characterized by the fact that the motive of action is not clear, and even though it may be harmful to oneself and others, it repetitively suppresses the impulse. In addition, several studies have reported that IGD causes psychological problems such as depression and anxiety. Some researchers have conducted studies to diagnose Internet game addiction through a physiological mechanism study. According to these results, it has been reported that internet addiction is associated with emotional and behavioral problems related to autonomic nervous system function such as anxiety, depression, stress, and impulsivity. Nevertheless, IGD is not yet clearly defined. Recently, some of researches have been conducted on the characteristics of IGD that addiction to online games leads to the executive control decrease especially when individuals are exposed to game-related cues. Executive control is associated with the autonomic nervous system, which is related to vagus nerves. For this reason, some researchers tried to investigate internet addiction through autonomic nervous system analysis. The autonomic nervous system is a nervous system that maintains homeostasis by controlling important functions of the body with sensing and reacting to changes in internal and external stimuli of the human body. Analyzing the changes of the autonomic nervous system function enable to predict various diseases including cardiovascular diseases. Heart rate variability(HRV) is one of the non-invasive methods that can quantitatively analyze the sympathetic (SNS) and parasympathetic nerves(PNS) of the autonomic nervous system. HRV is a noninvasive electrocardiographic marker that reflects activities of the sympathetic and parasympathetic nerves of the autonomic nervous system. In general, the change between heartbeats is larger and more complex in the stable state, and regular and constant in the case of movement or stress. In general, the response of the sympathetic nervous system rapidly decreases the heart rate, and the response of the parasympathetic nervous system increases the heart rate slowly. Previous studies have reported that the autonomic nervous system changes in the IGD group and the control group are different. However, the autonomic nervous system is a biomarker that can be changed not only by IGD but also by cardiovascular disease and mental disorder. Also, there is a limitation of the autonomic nervous system change to the Internet game stimulation because there are individual differences in the Internet game stimulation. In this study, we assumed that the HRV parameter that can analyze the vagus nerve activity during game is suppressed by executive control dysfunction of IGD group. In addition, we analyzed the HRV changes according to the game stimulus, and developed the IGD level prediction algorithm using the deep learning model. As a result, IGD group showed the suppressed HRV high frequency(HF) parameter on game stimulation, and it would be due to the reduction of vagus nerve activity by executive control dysfunction of IGD group. To verify these results, we analyzed the gray meter volume (GMV) of dorsolateral prefrontal cortex (DLPFC) and ventrolateral prefrontal cortex (VLPFC) associated with executive control through structural MRI analysis of IGD group. As a result, the significance of the HF parameter, DLPFC and VLPFC of the IGD group was confirmed. Based on these results, we developed a deep learning model with inputs for HRV Time-Frequency domain and finally we can estimation four levels of (normal, mild, moderate, severe) IGD severity using HRV bio-signals.
URI
http://dcollection.hanyang.ac.kr/common/orgView/000000107155http://repository.hanyang.ac.kr/handle/20.500.11754/76071
Appears in Collections:
GRADUATE SCHOOL OF BIOMEDICAL SCIENCE AND ENGINEERING[S](의생명공학전문대학원) > BIOMEDICAL ENGINEERING(생체의공학과) > Theses (Ph.D.)
Files in This Item:
There are no files associated with this item.
Export
RIS (EndNote)
XLS (Excel)
XML


qrcode

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

BROWSE