帕金森病步态的可穿戴多源数据采集与特征识别方法

谢俊枭; 赵欢; 曹军义; 曹红梅; 屈秋民; 廖维新

doi:10.7652/xjtuxb202306004

您当前的位置：

首页 >

文章列表页 >

帕金森病步态的可穿戴多源数据采集与特征识别方法

专题医工交叉与人机交互 | 更新时间：2025-05-08

- 帕金森病步态的可穿戴多源数据采集与特征识别方法
- Multi-Source Gait Data Acquisition and Feature Recognition Method Based on Wearable Device for Parkinson's Disease
- 西安交通大学学报 2023年57卷第6期页码：29-38
- 作者机构：
  
  1. 西安交通大学机械工程学院,西安,710049
  2. 西安交通大学第一附属医院神经内科,西安,710061
  3. 香港中文大学机械与自动化工程学系,香港,999077
- 作者简介：
  
  : 2022-11-25。作者简介: 谢俊枭(2000—),男,博士生
  曹军义(通信作者),男,教授,博士生导师。基金项目: 国家重点研发计划资助项目(2021YFE0203400)
- 基金信息：
  
  : 2022-11-25。作者简介: 谢俊枭(2000—),男,博士生;曹军义(通信作者),男,教授,博士生导师。基金项目: 国家重点研发计划资助项目(2021YFE0203400);内地与香港联合资助计划资助项目(MHP/043/20)。
- DOI：10.7652/xjtuxb202306004
  中图分类号： TH823;TP274
- 网络首发：2023-06-10，
  
  纸质出版：2023
- 稿件说明：
移动端阅览
谢俊枭, 赵欢, 曹军义, 等. 帕金森病步态的可穿戴多源数据采集与特征识别方法[J]. 西安交通大学学报, 2023,57(6):29-38.

XIE Junxiao, ZHAO Huan, CAO Junyi, et al. Multi-Source Gait Data Acquisition and Feature Recognition Method Based on Wearable Device for Parkinson's Disease[J]. 2023, 57(6): 29-38.
谢俊枭, 赵欢, 曹军义, 等. 帕金森病步态的可穿戴多源数据采集与特征识别方法[J]. 西安交通大学学报, 2023,57(6):29-38. DOI： 10.7652/xjtuxb202306004.

XIE Junxiao, ZHAO Huan, CAO Junyi, et al. Multi-Source Gait Data Acquisition and Feature Recognition Method Based on Wearable Device for Parkinson's Disease[J]. 2023, 57(6): 29-38. DOI： 10.7652/xjtuxb202306004.

摘要

针对帕金森病患者因早期症状隐匿、个体差异显著而导致该病早期识别与健康量化表征困难的问题

提出了一种柔性可穿戴的多源步态数据采集与特征分类方法。开发了柔性可穿戴步态采集系统

通过融合16通道柔性压阻传感器与8通道柔性压电传感器分别采集足底压力和形变信息

并针对各传感器设计了多源信号调理和传输电路

实现了多源步态信号的实时采集与无线传输。利用所设计的系统采集了帕金森患者与健康对照组的正常行走数据

通过对步态数据进行特征构建和显著性差异分析

提出了帕金森病特征识别方法

获取了具有识别效果的显著差异性特征。研究结果表明:帕金森病患者的时相特征不对称度、峰值变异系数、步态周期变异系数等特征值明显大于健康对照组。以支持向量机和k-近邻作为分类器分别对比了单一数据和多源数据特征用于帕金森病诊断的结果

发现当同时使用2种信号特征时

分类准确率比单一数据特征高6%～10%

其中基于k-近邻算法的帕金森病分类准确率达到89.66%。所提方法能有效提高帕金森病识别率

为帕金森病步态异常量化分析与诊断的临床应用奠定了基础。

Abstract

As early symptoms of Parkinson's disease patients are hidden

and distinct individual differences make it more difficult to identify early gait abnormalities and characterize healthy states quantitatively

a multi-source gait data acquisition and feature recognition method based on a wearable device was proposed in this paper. The multi-source gait data acquisition system can collect plantar pressure and deformation by fusing a 16-channel flexible piezoresistive sensor and an 8-channel flexible piezoelectric sensor. Through the design of the signal conditioning circuit

the real-time acquisition and wireless transmission of multi-source signals were achieved. Based on the designed system

the gait data of Parkinson's disease patients and healthy controls were collected. A method of feature recognition for Parkinson's disease was proposed based on distinctive features extracted by feature construction and significant difference analysis of gait data. The results showed that the features of Parkinson's disease patients

including asymmetry of phase features

coefficient of variation of peak value

and coefficient of variation of the gait cycle

were more distinctive than those of healthy controls. In addition

the Support Vector Machine and k-nearest neighbor were employed as classifiers to compare the accuracy between single and multi-source data features for diagnosing Parkinson's disease. The classification accuracy of multi-source gait signals was 6% to 10% higher than that of a single signal

in which the k-nearest neighbor classification accuracy of Parkinson's disease reached 89.66%. More importantly

the results demonstrated that the proposed method can effectively improve the ability to recognize Parkinson's disease and lay a foundation for quantitative analysis and diagnosis of Parkinson's disease in clinical application.

关键词

Keywords

references

ZHANG Huanghe, GUO Yi, ZANOTTO D. Accurate ambulatory gait analysis in walking and running using machine learning models [J]. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2020, 28(1): 191-202.[2] YUN J. User identification using gait patterns on UbiFloorII [J]. Sensors, 2011, 11(3): 2611-2639.[3] 周丙涛, 陈世强, 程宇阳, 等. 基于足底压力传感器与深度学习的生物身份识别 [J]. 仪器仪表学报, 2021, 42(7): 108-115.ZHOU Bingtao, CHEN Shiqiang, CHENG Yuyang, et al. Biometric identification based on plantar pressure sensor and deep learning [J]. Chinese Journal of Scientific Instrument,2021, 42(7): 108-115.[4] PARK J S, KOO S M, KIM C H. Classification of standing and walking states using ground reaction forces [J]. Sensors, 2021, 21(6): 2145.[5] 段鹏松, 周志一, 王超, 等. WiNet:一种适用于无线感知场景的步态识别模型 [J]. 西安交通大学学报, 2020, 54(7): 187-195.DUAN Pengsong, ZHOU Zhiyi, WANG Chao, et al. WiNet: a gait recognition model suitable for wireless sensing scene [J]. Journal of Xi'an Jiaotong University, 2020, 54(7): 187-195.[6] PEREIRA DOS SANTOS D M, FERREIRA NETO M, LEMOS M R, et al. Wearable system for early identification of Parkinson's disease symptoms through the evaluation of the gait training [C]//2019 IEEE 9th International Conference on Consumer Electronics(ICCE-Berlin). Piscataway, NJ, USA: IEEE, 2019: 51-56.[7] ZHAO Huan, WANG Ruixue, LEI Yaguo, et al. Severity level diagnosis of Parkinson's disease by ensemble K-nearest neighbor under imbalanced data [J]. Expert Systems with Applications, 2022, 189: 116113.[8] JANKOVIC J. Parkinson's disease: clinical features and diagnosis [J]. Journal of Neurology, Neurosurgery & Psychiatry, 2008, 79(4): 368-376.[9] IVKOVIC V, KURZ M J. Parkinson's disease influences the structural variations present in the leg swing kinematics [J]. Motor Control, 2011, 15(3): 359-375.[10] STEBBINS G T, GOETZ C G, BURN D J, et al. How to identify tremor dominant and postural instability/gait difficulty groups with the movement disorder society unified Parkinson's disease rating scale: comparison with the unified Parkinson's disease rating scale [J]. Movement Disorders, 2013, 28(5): 668-670.[11] HOEHN M M, YAHR M D. Parkinsonism: onset, progression, and mortality [J]. Neurology, 1998, 50(2): 318.[12] GOETZ C G, TILLEY B C, SHAFTMAN S R, et al. Movement disorder society-sponsored revision of the unified Parkinson's disease rating scale(MDS-UPDRS): scale presentation and clinimetric testing results [J]. Movement Disorders, 2008, 23(15): 2129-2170.[13] AN Weizhi, YU Shiqi, MAKIHARA Y, et al. Performance evaluation of model-based gait on multi-view very large population database with pose sequences [J]. IEEE Transactions on Biometrics, Behavior, and Identity Science, 2020, 2(4): 421-430.[14] UDDIN M Z, MURAMATSU D, TAKEMURA N, et al. Spatio-temporal silhouette sequence reconstruction for gait recognition against occlusion [J]. IPSJ Transactions on Computer Vision and Applications, 2019, 11(1): 9.[15] VERLEKAR T T, SOARES L D, CORREIA P L. Gait recognition in the wild using shadow silhouettes [J]. Image and Vision Computing, 2018, 76: 1-13.[16] SILVA DE LIMA A L, EVERS L J W, HAHN T, et al. Freezing of gait and fall detection in Parkinson's disease using wearable sensors: a systematic review [J]. Journal of Neurology, 2017, 264(8): 1642-1654.[17] HUA Rui, WANG Ya. Monitoring insole(MONI): a low power solution toward daily gait monitoring and analysis [J]. IEEE Sensors Journal, 2019, 19(15): 6410-6420.[18] POPOVIC M B, DJURIC-JOVICIC M, RADOVANOVIC S, et al. A simple method to assess freezing of gait in Parkinson's disease patients [J]. Brazilian Journal of Medical and Biological Research, 2010, 43(9): 883-889.[19] 李波陈. 冻结步态可穿戴监测方法研究 [D]. 合肥: 中国科学技术大学, 2022.[20] SVEINBJORNSDOTTIR S. The clinical symptoms of Parkinson's disease [J]. Journal of Neurochemistry, 2016, 139(S1): 318-324.[21] WANG Wei, CAO Junyi, YU Jian, et al. Self-powered smart insole for monitoring human gait signals [J]. Sensors, 2019, 19(24): 5336.[22] 孙东杰, 宋爱国. 基于传感阵列的动态足底压力分布测量系统 [J]. 仪器仪表学报, 2022, 43(6): 83-91.SUN Dongjie, SONG Aiguo. A dynamic plantar pressure distribution measurement system based on sensor array [J]. Chinese Journal of Scientific Instrument, 2022, 43(6): 83-91.[23] ZHAO Huan, WANG Ruixue, QI Dexin, et al. Wearable gait monitoring for diagnosis of neurodegenerative diseases [J]. Measurement, 2022, 202: 111839.[24] ESPAY A J, BONATO P, NAHAB F B, et al. Technology in Parkinson's disease: challenges and opportunities [J]. Movement Disorders, 2016, 31(9): 1272-1282.[25] CHEN Shanshan, LACH J, LO B, et al. Toward pervasive gait analysis with wearable sensors: a systematic review [J]. IEEE Journal of Biomedical and Health Informatics, 2016, 20(6): 1521-1537.[26] 具典淑, 周智, 欧进萍. PVDF压电薄膜的应变传感特性研究 [J]. 功能材料, 2004, 35(4): 450-452, 456.JU Dianshu, ZHOU Zhi, OU Jinping. Study on strain-sensing of PVDF films [J]. Journal of Functional Materials, 2004, 35(4): 450-452, 456.[27] KOUR N, GUPTA S, ARORA S. Sensor technology with gait as a diagnostic tool for assessment of Parkinson's disease: a survey [J]. Multimedia Tools and Applications, 2022: 1-37.[28] 刘俊杰. 基于机器学习与步态分析的帕金森病患者定量评估与分类方法研究 [D]. 杭州: 杭州电子科技大学, 2021.[29] VEERARAGAVAN S, GOPALAI A A, GOUWANDA D, et al. Parkinson's disease diagnosis and severity assessment using ground reaction forces and neural networks [J]. Frontiers in Physiology, 2020, 11: 587057.[30] ABDULHAY E, ARUNKUMAR N, NARASIMHAN K, et al. Gait and tremor investigation using machine learning techniques for the diagnosis of Parkinson disease [J]. Future Generation Computer Systems, 2018, 83: 366-373.[31] YUNGHER D A, MORRIS T R, DILDA V, et al. Temporal characteristics of high-frequency lower-limb oscillation during freezing of gait in Parkinson's disease [J]. Parkinson's Disease, 2014, 2014: 606427.[32] 余建, 曹军义, 王伟, 等. 人体步态复杂度的递归图和递归定量分析研究 [J]. 西安交通大学学报, 2017, 51(10): 47-52, 70.YU Jian, CAO Junyi, WANG Wei, et al. Recurrence plot and recurrence quantification analysis of human gait complexity [J]. Journal of Xi'an Jiaotong University, 2017, 51(10): 47-52, 70.[33] ZHAO Huan, CAO Junyi, WANG Ruixue, et al. Accurate identification of Parkinson's disease by distinctive features and ensemble decision trees [J]. Biomedical Signal Processing and Control, 2021, 69: 102860.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

水蒸气掺混对氨裂解气旋流燃烧与排放特性的影响研究

语义区分和子图整合驱动的知识图谱表示学习方法

电子级超纯氮深冷空分流程节能优化研究

采用机器学习的大规模并行程序I/O性能建模与预测方法

超临界二氧化碳向心透平轮背间隙内风阻损失及流动特性研究