Cited 0 time in webofscience Cited 0 time in scopus

Subject-specific real time motor imaginary detection scheme for robot-aided hand rehabilitation

Subject-specific real time motor imaginary detection scheme for robot-aided hand rehabilitation
Translated Title
로봇을 이용한 손 재활치료를 위한 피험자 맞춤형 실시간 움직임 의도 파악 방법
Chun, Rae Chang
DGIST Authors
Chun, Rae Chang; Kim, Jong Hyun
Kim, Jong Hyun
Jang, Sung Ho
Issue Date
Available Date
Degree Date
2016. 2
Human intentionMotor imageryVoluntary movementEvent-related Desynchronization(ERD)Electroencephalogram(EEG)Support Vector Machine(SVM)Machine learningReal time signal processing움직임 의도인간의 의도자발적인 움직임사건관련 비동기화뇌전도서포트 벡터 머신기계학습실시간 신호처리
This study is a motor imaginary detection scheme for rehabilitation. Recently, detecting motor imaginary movement based on brain mapping device has been applied to improve robot-aided therapy for rehabilitation. Our goal is to develop a simple method that perform a system in real time to make a natural movement, to build subject-specific real time code to realize a system that help subject-specific rehabilitation therapy and classify ERD and Fake MI and eliminate Fake MI for correct rehabilitation therapy. We opt to EEG for brain imaging modality and using Matlab software for EEG signal processing. Event-related desynchronization (ERD) occurs in specific frequency band in brain wave when human has intention of movement. To detect ERD, in this thesis, we utilize a method called Machine Learning. The machine learning algorithm we applied in this study was Support Vector Machine (SVM). The result of SVM represents low success trial of ERD and low false detection of Fake MI. The algorithm that remove Fake MI also eliminate the ERD and that cause low success trials. We built the real time system that be able to perform voluntary-like movement. During building the system, we have found problems like Fake MI. Fake MI is the factor that interrupt a correct rehabilitation therapy. ⓒ 2016 DGIST
Table Of Contents
Ⅰ. INTRODUCTION 1-- 1.1 Background 1-- 1.2 Contribution Point 4-- 1.3 Purpose 5-- Ⅱ. METHOD 6-- 2.1 Participants 6-- 2.2 Device 6-- 2.3 Experimental Protocol 9-- 2.3.1 Surroundings 9-- 2.3.2 Experiment setup and procedure 10-- 2.4 System Configuration 13-- 2.4.1 Signal Processing 14-- 2.4.2 Machine Learning 16-- 2.4.3 ERD / Fake MI classification 20-- Ⅲ. RESULT 29-- 3.1 EEG Experiment Result 29-- 3.2 Machine Learning Result 32-- Ⅳ. DISCUSSION 35-- Ⅴ. CONCLUSION 37-- Ⅵ. APPENDIX 38-- Reference 47-- --
Robotics Engineering
Related Researcher
  • Author Kim, Jonghyun REL(Rehabilitation Engineering Laboratory)
  • Research Interests Rehabilitation engineering; Robotics; Nonlinear control
Department of Robotics EngineeringThesesMaster

qrcode mendeley

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