中国科学院机构知识库网格
Chinese Academy of Sciences Institutional Repositories Grid
首页 机构 成果 学者
  
  1. CAS IR Grid
  2. 自动化研究所
  3. 中国科学院自动化研究所
  4. 复杂系统认知与决策实验室
  5. 先进机器人控制团队
An Automatic Rehabilitation Assessment System for Hand Function Based on Leap Motion and Ensemble Learning

文献类型：期刊论文

作者Li, Chenguang1,2; Cheng, Long1,2; Yang, Hongjun1,2; Zou, Yongxiang1,2; Huang, Fubiao3
刊名CYBERNETICS AND SYSTEMS
出版日期2020-10-01
页码23
关键词Automatic assessment Brunnstrom assessment ensemble learning Fugl-Meyer assessment hand function assessment rehabilitation Leap Motion
ISSN号0196-9722
DOI10.1080/01969722.2020.1827798
通讯作者Cheng, Long(long.cheng@ia.ac.cn)
英文摘要For stroke patients, hand function assessment is an important part of the hand rehabilitation process. The hand function assessment, however, requires the patient to complete a series of actions under the guidance of the therapist who then scores the patient's performance. This type of assessment is both time-consuming and highly subjective. Therefore, in order to achieve a fast, objective and accurate assessment, this paper adopts a non-contact infrared imaging device, Leap Motion, to measure the patient's motion information and then uses these motion information to infer the hand's rehabilitation level. This paper improves the traditional way of hand function assessment from the following aspects. Only three coherent movements (finger opposition, lift wrist and stretch fingers) are required to complete the assessment, which makes the assessment time shorter and the assessment process easier. At the same time, an assessment algorithm based on the Ensemble Learning is proposed and integrated into the automatic hand function assessment system. In addition, the virtual reality game has been implemented in the assessment system to ensure a satisfactory interaction with patients, which makes the assessment process more interesting and convenient. Using this system, 50 stroke patients underwent clinical trials with the Brunnstrom and Fugl-Meyer assessment scales. The matching rate between the automatic assessment result and the manual Brunnstrom assessment result is 92%, while the matching rate with the Fugl-Meyer assessment result is 82%. Furthermore, Wilcoxon Signed-Rank test and Kappa test are also used to validate the consistency between the automatic assessment results and the manual assessment results. These experiments illustrate that this automatic assessment system is fast, comfortable and reliable.
WOS关键词SIGN-LANGUAGE ; RECOGNITION ; CONTROLLER ; ACCURACY
资助项目Beijing Municipal Natural Science Foundation[JQ19020] ; Beijing Municipal Natural Science Foundation[L182060] ; National Natural Science Foundation of China[U1913209] ; National Natural Science Foundation of China[61873268]
WOS研究方向Computer Science
语种英语
WOS记录号WOS:000578283700001
出版者TAYLOR & FRANCIS INC
资助机构Beijing Municipal Natural Science Foundation ; National Natural Science Foundation of China
源URL[http://ir.ia.ac.cn/handle/173211/42131]  
专题自动化研究所_复杂系统管理与控制国家重点实验室_先进机器人控制团队
通讯作者Cheng, Long
作者单位1.Univ Chinese Acad Sci, Sch Artificial Intelligence, Beijing, Peoples R China
2.Chinese Acad Sci, Inst Automat, State Key Lab Control & Management Complex Syst, Beijing, Peoples R China
3.China Rehabil Res Ctr, Dept Occupat Therapy, Beijing, Peoples R China
推荐引用方式
GB/T 7714		 Li, Chenguang,Cheng, Long,Yang, Hongjun,et al. An Automatic Rehabilitation Assessment System for Hand Function Based on Leap Motion and Ensemble Learning[J]. CYBERNETICS AND SYSTEMS,2020:23.
APA Li, Chenguang,Cheng, Long,Yang, Hongjun,Zou, Yongxiang,&Huang, Fubiao.(2020).An Automatic Rehabilitation Assessment System for Hand Function Based on Leap Motion and Ensemble Learning.CYBERNETICS AND SYSTEMS,23.
MLA Li, Chenguang,et al."An Automatic Rehabilitation Assessment System for Hand Function Based on Leap Motion and Ensemble Learning".CYBERNETICS AND SYSTEMS (2020):23.

入库方式： OAI收割

来源：自动化研究所

浏览0
下载0
收藏0
其他版本

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。