基于Xception-LSTM的下肢运动能力评价方法

doi:10.3969/j.issn.1006-9771.2020.06.004

《中国康复理论与实践》 ›› 2020, Vol. 26 ›› Issue (6): 643-647.doi: 10.3969/j.issn.1006-9771.2020.06.004

• 专题康复体育和运动功能康复研究 • 上一篇下一篇

基于Xception-LSTM的下肢运动能力评价方法

张燕¹,王铭玥¹,王婕¹(),姜恺宁¹,张筠晗²

1.河北工业大学人工智能与数据科学学院,天津市 300130
2.郑州大学国际学院,河南郑州市 450001

收稿日期:2019-07-04 修回日期:2019-08-28 出版日期:2020-06-25 发布日期:2020-06-29
通讯作者: 王婕 E-mail:wangjie@hebut.edu.cn
作者简介:张燕(1975-),女,汉族,河北石家庄市人,博士,教授,博士研究生导师,主要研究方向：智能康复辅具与模式识别。
基金资助:
1.国家自然科学基金项目(61773151);2.河北省自然科学基金项目(F2018202279)

An Evaluation of Lower Limb Motor Ability Based on Xception-LSTM

ZHANG Yan¹,WANG Ming-yue¹,WANG Jie¹(),JIANG Kai-ning¹,ZHANG Jun-han²

1. School of Artificial Intelligence and Data Science, Hebei University of Technology, Tianjin 300130, China
2. International College, Zhengzhou University, Zhengzhou, Henan 450001, China

Received:2019-07-04 Revised:2019-08-28 Published:2020-06-25 Online:2020-06-29
Contact: WANG Jie E-mail:wangjie@hebut.edu.cn
Supported by:
National Natural Science Foundation of China(61773151);Natural Science Foundation of Hebei(F2018202279)

摘要/Abstract

摘要：

目的建立下肢运动功能评估的算法。方法 2016年8月至2017年3月,40例受试者分为年轻健康组(n = 20)、中年组(n = 10)和老年组(n = 10)。采集受试者的步态视频、膝关节角度和地面反作用力,并利用ViBe算法对步态视频进行步态轮廓提取;利用Xception-LSTM网络提取步态图像特征,并与膝关节角度和地面反作用力在特征层进行融合;将融合特征经核主成分分析降维处理,生成步行能力评分(GAS),并对GAS和威斯康辛步态量表(WGS)评分进行相关性分析。结果中年组和老年组的GAS均明显低于年轻健康组(t > 4.164, P < 0.01),且老年组明显低于中年组( t = 7.338, P < 0.01)。GAS与WGS评分呈负相关( r = -0.91, P < 0.01)。结论 GAS能够量化评估下肢运动能力,可帮助制定康复方案,适配助行设备。

关键词: 老年人, 下肢, 运动能力, 迁移学习, 卷积神经网络, 循环神经网络, 特征提取

Abstract:

Objective To establish an algorithm to quantitatively evaluate the lower limb motor ability.Methods From August, 2016 to March, 2017, 40 subjects were divided into young healthy group ( n = 20), middle-aged group (n = 10) and elderly group (n = 10). The gait video, knee angle and ground reaction force of the subjects were collected, and the gait contour was extracted from the gait video by using the ViBe algorithm. The gait image feature was extracted by Xception-LSTM, and fused it with the knee joint angle and the ground reaction force in the feature layer. The fusion features were reduced in dimension by kernel principal component analysis, and the gait ability score (GAS) was established. All the subjects were assessed with Wisconsin Gait Scale (WGS).Results GAS was less in middle-aged group and elderly group than in the young healthy group (t > 4.164, P < 0.01), and was less in the elderly group than in the middle-aged group ( t = 7.338, P < 0.01). GAS was negative correlated with the score of WGS ( r = -0.91, P < 0.01). Conclusion The lower limb exercise ability could be quantified with GAS, which may be applied in developing rehabilitation and fitting walking aids.

Key words: elderly, lower limb, motor ability, transfer learning, convolutional neural network, recurrent neural network, feature extraction

中图分类号:

R331

张燕,王铭玥,王婕,姜恺宁,张筠晗. 基于Xception-LSTM的下肢运动能力评价方法[J]. 《中国康复理论与实践》, 2020, 26(6): 643-647.

ZHANG Yan,WANG Ming-yue,WANG Jie,JIANG Kai-ning,ZHANG Jun-han. An Evaluation of Lower Limb Motor Ability Based on Xception-LSTM[J]. 《Chinese Journal of Rehabilitation Theory and Practice》, 2020, 26(6): 643-647.

图/表 4

参考文献 27

[1]	原新. 积极应对人口老龄化是新时代的国家战略[J]. 人口研究, 2018, 42(3):3-8.
[2]	顾琳燕, 邱华平, 张晟宇, 等. 正常人群步行足偏角与年龄特征分析[J]. 生物医学工程学杂志, 2018, 35(1):45-48.
[3]	王娜, 瓮长水, 朱才兴, 等. 老年人下肢肌力、协调性和反应时的年龄特征相关性[J]. 中国康复理论与实践, 2011, 17(12):1155-1157.
[4]	高亚南, 许永利, 陈雪丽. 老年综合评估在老年康复中的应用[J]. 中国康复理论与实践, 2013, 19(5):452-456.
[5]	Prodinger B, Cieza A, Oberhauser C, et al. Toward the International Classification of Functioning, Disability and Health (ICF) rehabilitation set: a minimal generic set of domains for rehabilitation as a health strategy[J]. Arch Phys Med Rehabil, 2016, 97(6):875-884. doi: 10.1016/j.apmr.2015.12.030
[6]	Bartoszek G, Fischer U, von Clarenau S C, et al. Development of an International Classification of Functioning, Disability and Health (ICF)-based standard set to describe the impact of joint contractures on participation of older individuals in geriatric care settings[J]. Arch Gerontol Geriatr, 2015, 61(1):61-66. doi: 10.1016/j.archger.2015.03.005
[7]	Gabler G, Coenen M, Lycett D, et al. Towards a standardized nutrition and dietetics terminology for clinical practice: an Austrian multicenter clinical documentation analysis based on the International Classification of Functioning, Disability and Health (ICF)-Dietetics[J]. Clin Nutr, 2019, 38(2):791-799. doi: 10.1016/j.clnu.2018.02.031
[8]	Morris J N, Howard E P, Fries B E, et al. Using the community health assessment to screen for continued driving[J]. Accid Anal Prev, 2014, 63:104-110. doi: 10.1016/j.aap.2013.10.030
[9]	Wada T, Ishine M, Sakagami T, et al. Depression, activities of daily living, and quality of life of community-dwelling elderly in three Asian countries: Indonesia, Vietnam, and Japan[J]. Arch Gerontol Geriatr, 2005, 41(3):271-280. doi: 10.1016/j.archger.2005.03.003
[10]	McMulkin M L, MacWilliams B A. Application of the Gillette Gait Index, Gait Deviation Index and Gait Profile Score to multiple clinical pediatric populations[J]. Gait Posture, 2015, 41(2):608-612. doi: 10.1016/j.gaitpost.2015.01.005 pmid: 25623856
[11]	Balasubramanian C K, Clark D J, Gouelle A. Validity of the Gait Variability Index in older adults: effect of aging and mobility impairments[J]. Gait Posture, 2015, 41(4):941-946. doi: 10.1016/j.gaitpost.2015.03.349 pmid: 25882115
[12]	Christian J, Kroll J, Schwameder H. Comparison of the Classifier Oriented Gait Score and the Gait Profile Score based on imitated gait impairments[J]. Gait Posture, 2017, 55:49-54. doi: S0966-6362(17)30119-4 pmid: 28411445
[13]	Kwon J W, Son S M, Lee N K. Changes of kinematic parameters of lower extremities with gait speed: a 3D motion analysis study[J]. J Phys Ther Sci, 2015, 27(2):477-479. doi: 10.1589/jpts.27.477
[14]	Cao J, Xie S Q, Das R, et al. Control strategies for effective robot assisted gait rehabilitation: the state of art and future prospects[J]. Med Eng Phys, 2014, 36(12):1555-1566. doi: 10.1016/j.medengphy.2014.08.005
[15]	Mehdizadeh S. The largest Lyapunov exponent of gait in young and elderly individuals: a systematic review[J]. Gait Posture, 2018, 60(2):241-250. doi: 10.1016/j.gaitpost.2017.12.016
[16]	刘程程, 元香南, 张立新, 等. 老年人与健康大学生平地行走时足底压力特征比较[J]. 中国康复理论与实践, 2015, 21(5):544-548.
[17]	Osoba M Y, Rao A K, Agrawal S K, et al. Balance and gait in the elderly: a contemporary review[J]. Laryngoscope Investig Otolaryngol, 2019, 4(1):143-153.
[18]	张庆来, 张林. 老年人步态移动稳定性的增龄特征研究[J]. 中国运动医学杂志, 2017, 36(7):599-604.
[19]	周飞燕, 金林鹏, 董军. 卷积神经网络研究综述[J]. 计算机学报, 2017, 40(6):1229-1251.
[20]	常亮, 邓小明, 周明全, 等. 图像理解中的卷积神经网络[J]. 自动化学报, 2016, 42(9):1300-1312.
[21]	Mamoshina P, Vieira A, Putin E, et al. Applications of deep learning in biomedicine[J]. Mol Pharm, 2016, 13(5):1445-1454. doi: 10.1021/acs.molpharmaceut.5b00982
[22]	余镇, 吴凌云, 倪东, 等. 基于深度学习的胎儿颜面部超声标准切面自动识别[J]. 中国生物医学工程学报, 2017, 36(3):267-275.
[23]	梁翠霞, 李明强, 边兆英, 等. 基于深度学习特征的乳腺肿瘤分类模型评估[J]. 南方医科大学学报, 2019, 39(1):88-92.
[24]	刘天亮, 谯庆伟, 万俊伟, 等. 融合空间-时间双网络流和视觉注意的人体行为识别[J]. 电子与信息学报, 2018, 40(10):2395-2401.
[25]	Yu Y, Si X, Hu C, et al. A review of recurrent neural networks: LSTM cells and network architectures[J]. Neural Comput, 2019, 31(7):1235-1270. doi: 10.1162/neco_a_01199
[26]	Ergen T, Kozat S S. Efficient online learning algorithms based on LSTM neural networks[J]. IEEE Trans Neural Netw Learn Syst, 2017, 29(8):3772-3783. doi: 10.1109/TNNLS.2017.2741598
[27]	Tsironi E, Barros P, Weber C, et al. An analysis of convolutional long short-term memory recurrent neural networks for gesture recognition[J]. Neurocomputing, 2017, 268:76-86. doi: 10.1016/j.neucom.2016.12.088

基于Xception-LSTM的下肢运动能力评价方法

An Evaluation of Lower Limb Motor Ability Based on Xception-LSTM

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 4

参考文献 27

相关文章 15

Metrics

本文评价

推荐阅读 0

组别	n	GAS
年轻健康组	20	100.576±14.331
中年组	10	78.214±8.045^a
老年组	10	50.823±10.046^b,c

[1]	王航宇, 葛可可, 范永红, 都丽露, 邹敏, 封磊. 基于ICD-11和ICF主动式音乐疗法改善认知障碍老年人认知功能的系统综述[J]. 《中国康复理论与实践》, 2024, 30(1): 36-43.
[2]	葛可可, 范永红, 王航宇, 都丽露, 李长江, 邹敏. 失眠老年人正念干预健康效益的系统综述[J]. 《中国康复理论与实践》, 2024, 30(1): 54-60.
[3]	杨雅楠, 穆丽萍, 邢凤梅, 薛新宏, 王晓光, 陶徉聿, 孙竹梅, 张小丽. 基于计划行为理论的干预对肌少症老年人肌肉衰减状况及平衡能力的效果[J]. 《中国康复理论与实践》, 2023, 29(8): 869-874.
[4]	王亚楠, 刘西花. 脑卒中偏瘫患者主观和客观平衡功能测量的相关性及预测效能[J]. 《中国康复理论与实践》, 2023, 29(8): 890-895.
[5]	蒋长好, 黄辰, 高晓妍, 戴元富, 赵国明. 神经反馈训练对老年人认知功能效果的系统综述[J]. 《中国康复理论与实践》, 2023, 29(8): 903-909.
[6]	刘辉, 尹航, 贾绍辉, 邱服冰. 适用于残疾儿童青少年运动功能和运动能力测量工具的结构、内容和心理测量特性的系统综述[J]. 《中国康复理论与实践》, 2023, 29(6): 630-638.
[7]	王一吉, 周红俊, 何泽佳, 刘根林, 郑樱, 郝春霞, 卫波, 康海琼, 张缨, 逯晓蕾, 袁媛, 蒙倩茹. 不完全性脊髓损伤患者运动功能对称性与步态对称性的关系[J]. 《中国康复理论与实践》, 2023, 29(6): 639-645.
[8]	谭晓欢, 姜桂萍, 黄芯怡, 王丹丹, 张磊, 卜庆国, 吴雪萍. 社区高龄老年人下肢肌力与步速的关系：柔韧和动态平衡的链式中介作用[J]. 《中国康复理论与实践》, 2023, 29(6): 646-653.
[9]	张倩, 孙新亭. 视频镜像训练对脑卒中恢复期患者下肢功能的效果[J]. 《中国康复理论与实践》, 2023, 29(6): 703-707.
[10]	赵雅娴, 唐芷晴, 孙新亭, 王荣荣, 刘天昊, 张皓. 不同强度穿戴式下肢康复机器人训练对脑卒中后下肢功能的效果[J]. 《中国康复理论与实践》, 2023, 29(5): 497-503.
[11]	金振华, 陈玲, 刘勇. 基于自我效能理论的数字化步行功能训练对脑卒中患者下肢功能的效果[J]. 《中国康复理论与实践》, 2023, 29(5): 504-509.
[12]	袁媛, 杨剑. 社区老年人身体活动融合慢性病管理的健康效益：Scoping综述[J]. 《中国康复理论与实践》, 2023, 29(5): 541-550.
[13]	潘海鸿, 李泰薇, 刘远才, 卢美铭, 陈琳. 简易滑动式腿部康复设备机构设计与分析[J]. 《中国康复理论与实践》, 2023, 29(4): 396-401.
[14]	黄凯琪, 辛榕, 凌耿强, 王朴. 重复经颅磁刺激治疗脑卒中后下肢运动功能障碍的Scoping综述[J]. 《中国康复理论与实践》, 2023, 29(3): 286-293.
[15]	马启寿, 李中元, 符卫卫, 林茜. 重复经颅磁刺激对恢复期脑卒中患者步行的效果[J]. 《中国康复理论与实践》, 2023, 29(2): 167-173.