基於數據融合模型的機器學習 對甲基苯丙胺使用障礙的多生理訊號號分析;Multi-bio-signal analysis of Methamphetamine Use Disorder by Machine Learning on data fusion model

NCUIR > College of Electrical Engineering & Computer Science > Graduate Institute of Computer Science and Information Engineering > Electronic Thesis & Dissertation > Item 987654321/86802

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/86802

Title:	基於數據融合模型的機器學習對甲基苯丙胺使用障礙的多生理訊號號分析;Multi-bio-signal analysis of Methamphetamine Use Disorder by Machine Learning on data fusion model
Authors:	葉軍;Yeh, Chun
Contributors:	資訊工程學系
Keywords:	甲基苯丙胺;生物信號;心率變異性;心電圖;腦電圖;皮膚電反應;數據融合;多模態數據;虛擬現實;機器學習;Methamphetamine;Bio-signal;Heart rate variability;Electrocardiography;Electroencephalography;Galvanic Skin Response;Data Fusion;Multimodal Data;Virtual reality;Machine Learning
Date:	2021-09-07
Issue Date:	2021-12-07 13:14:19 (UTC+8)
Publisher:	國立中央大學
Abstract:	根據我們之前在 2020 年發表的研究，我們通過皮膚電反應 (GSR)、心率變異性 (HRV) 和腦電圖 (EEG) 等生物信號，成功建立了甲基苯丙胺使用障礙 (MUD) 患者的分類模型並成功分類他們。在以MUD患者為實驗組、健康人為對照組的參與者收集HRV、GSR、EEG等生物傳感器信號數據後，我們使用每個生物傳感器信號進行分類。參與過我們的VR系統後，在分類MUD與正常人的模型中，於 HRV 中，我們獲得了 80.01% 的準確率；於 GSR 中，我們獲得了 78.12% 的準確率；於 EEG 中，我們獲得了 92.30% 的準確率。在這項研究中，我們通過結合這三種類型的數據來提高準確性。結果，我們的多模態生物傳感器模型獲得了 99.01% 的準確率。有了這個虛擬現實系統和預測模型，我們能夠提供一個更有效的甲基苯丙胺治療系統。 ;According to our previous study published in 2020, We successfully established a classification model for patients with Methamphetamine Use Disorder (MUD) through biological signals such as Galvanic Skin Response (GSR), Heart Rate Variability (HRV) and Electroencephalogram (EEG) and successfully classified them in our VR system. After collecting bio-sensor signal data such as HRV, GSR, and EEG from participants with MUD patients as experiment group and healthy people as control group, we used each bio-sensor signal to classify. In the classification between MUD and healthy subjects after participating our VR system; in HRV, we got 80.01% accuracy. In GSR, we got 78.12% accuracy. And in EEG we got 92.30% accuracy. In this study, we recruited more participants and tried to improve the accuracy by combining these three types of data. As a result, we got 99.01% accuracy by our multimodal bio-sensor model. With this VR system and forecast model, we are able to provide a more effective system in Methamphetamine treatment.
Appears in Collections:	[Graduate Institute of Computer Science and Information Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	126	View/Open

社群 sharing

Loading...