SIA OpenIR  > 机器人学研究室
基于表面肌电的手部运动康复与量化评估技术研究
Alternative TitleResearch on Hand Movement Rehabilitation and Quantitative Evaluation Technology Based on Surface Electromyography
胡少康
Department机器人学研究室
Thesis Advisor赵新刚
Keyword手部康复机器人 表面肌电信号 人机交互 康复评估 机器学习
Pages82页
Degree Discipline模式识别与智能系统
Degree Name硕士
2021-05-21
Degree Grantor中国科学院沈阳自动化研究所
Place of Conferral沈阳
Abstract近年来,脑卒中已经成为全球负担最重的疾病,是致死和致残的最大原因之一,另外脑卒中患者数量仍在逐渐增加并趋向于年轻化。大多脑卒中患者存在手部运动功能障碍,极大影响患者的日常生活。现代康复理论和实践表明,科学的运动康复训练可以有效的帮助脑卒中患者恢复身体机能。但传统康复训练存在训练效率低和训练强度不足等问题,因此手部康复机器人成为更好的选择。本文针对传统康复过程中康复训练和康复评估效率低、康复成本高等问题,研究了基于表面肌电的脑卒中患者手部动作识别和量化康复评估方法,并基于这两项技术研发了基于表面肌电的手部康复机器人系统。本文具体研究内容如下: 首先,为了便于研究脑卒中患者手部动作识别和康复评估技术,建立了脑卒中患者的手部表面肌电数据集。该数据集包含43名脑卒中患者的9种常见手部康复动作的表面肌电数据,覆盖全部康复等级,适合进行患者手部动作识别和量化康复评估的研究。并介绍了表面肌电数据的特征工程,包括数据预处理、特征提取和特征选择。数据预处理主要指肌电数据滤波和归一化方法,特征提取和特征选择介绍了传统方法和自动方法,通过提出的自动特征提取和特征选择方法,可以得到适合任务的最优特征集。其次,针对脑卒中患者手部动作识别率低的问题,提出了一种基于表面肌电的脑卒中患者手部动作识别方法。通过对30名不同等级患者的表面肌电数据进行特征提取和特征选择,筛选出了最合适的滑动窗参数及特征,使用设计的分类模型进行动作识别,实验结果表明该方法性能优于常见分类模型,最后进行在线实验并分析了实时性。再次,针对传统康复评估方法的主观性强、实时性差和效率低等问题,提出了基于表面肌电的脑卒中患者手部量化康复评估方法。通过特征工程筛选出适合康复评估的特征,使用设计的量化评估模型进行康复评估,实验结果表明该方法优于其他模型,并确定了适合康复评估的康复动作。并探究了康复评估的应用难点,取得了一定的成果。 最后,基于脑卒中患者手部动作识别和量化康复评估两项技术,设计了基于表面肌电的手部康复机器人系统,具有康复训练、量化康复评估和康复方案自调整功能,适用于家庭和社区式的脑卒中患者康复训练。
Other AbstractIn recent years, stroke has become the most burdened disease in the world, and it is one of the biggest causes of death and disability. In addition, the number of stroke patients is still increasing and tends to be younger. Most stroke patients have hand movement dysfunction, which greatly affects the patients' daily life. Modern rehabilitation theory and practice show that scientific sports rehabilitation training is an effective way to restore the physical functions of stroke patients. However, traditional rehabilitation training has problems such as low training efficiency and insufficient training intensity, so hand rehabilitation robots have become a better choice. Aiming at the problems of low efficiency of rehabilitation training and rehabilitation evaluation and high cost of rehabilitation in the traditional rehabilitation process, this paper studies the hand movement recognition and quantitative rehabilitation evaluation methods of stroke patients based on surface electromyography, and based on these two technologies, develops the sEMG-based hand rehabilitation robot system.The specific research content of this article is as follows: Firstly, in order to facilitate the study of hand movement recognition and rehabilitation assessment techniques of stroke patients, a data set of surface electromyography of the hands of stroke patients was established. This data set contains surface EMG data of 9 common hand rehabilitation actions of 43 stroke patients, covering all rehabilitation levels, and is suitable for research on patient hand movement recognition and quantitative rehabilitation assessment. The feature engineering of surface EMG data is introduced, including data preprocessing, feature extraction and feature selection. Data preprocessing mainly refers to EMG data filtering and normalization methods. Feature extraction and feature selection introduce traditional methods and automatic methods. Through the proposed automatic feature extraction and feature selection methods, the optimal feature set suitable for the task can be obtained. Secondly, aiming at the problem of the low recognition rate of hand movements in stroke patients, a surface EMG-based hand movement recognition method in stroke patients was proposed. Through feature extraction and feature selection of the surface EMG data of 30 patients of different levels, the most suitable sliding window parameters and features are screened out, and the designed classification model is used for action recognition. The experimental results show that the performance of this method is better than common classification. Model, and finally conduct online experiments and analyze the real-time performance. Thirdly, aiming at the problems of strong subjectivity, poor real-time performance and low efficiency of traditional rehabilitation evaluation methods, a quantitative rehabilitation evaluation method for stroke patients' hands based on surface electromyography is proposed. Feature engineering is used to screen out the characteristics suitable for rehabilitation evaluation, and the designed quantitative evaluation model is used for rehabilitation evaluation. The experimental results show that this method is superior to other models, and the rehabilitation actions suitable for rehabilitation evaluation are determined. And explored the application difficulties of rehabilitation assessment, and achieved certain results. Finally, based on the two technologies of hand movement recognition and quantitative rehabilitation evaluation of stroke patients, a hand rehabilitation robot system based on surface electromyography is designed. It has rehabilitation training, quantitative rehabilitation evaluation and self-adjustment functions of rehabilitation programs, which is suitable for family and community-style rehabilitation training for stroke patients.
Language中文
Contribution Rank1
Document Type学位论文
Identifierhttp://ir.sia.cn/handle/173321/28953
Collection机器人学研究室
Affiliation中国科学院沈阳自动化研究所
Recommended Citation
GB/T 7714
胡少康. 基于表面肌电的手部运动康复与量化评估技术研究[D]. 沈阳. 中国科学院沈阳自动化研究所,2021.
Files in This Item:
File Name/Size DocType Version Access License
基于表面肌电的手部运动康复与量化评估技术(9175KB)学位论文 开放获取CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[胡少康]'s Articles
Baidu academic
Similar articles in Baidu academic
[胡少康]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[胡少康]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

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