Please wait a minute...
浙江大学学报(工学版)  2021, Vol. 55 Issue (4): 639-647    DOI: 10.3785/j.issn.1008-973X.2021.04.005
计算机技术、电信技术     
基于腕部姿态的帕金森病用药后开-关期检测
张腾1,2,3(),蒋鑫龙1,2,3,*(),陈益强1,2,3,陈前1,2,3,米涛免4,陈彪4
1. 中国科学院 计算技术研究所,北京 100190
2. 移动计算与新型终端北京市重点实验室,北京 100190
3. 中国科学院大学计算机科学与技术学院,北京 100049
4. 首都医科大学宣武医院国家老年疾病临床研究中心,北京 100053
Wrist attitude-based Parkinson's disease ON/OFF state assessment after medication
Teng ZHANG1,2,3(),Xin-long JIANG1,2,3,*(),Yi-qiang CHEN1,2,3,Qian CHEN1,2,3,Tao-mian MI4,Piu CHAN4
1. Institute of Computing Technology, Chinese Academy of Sciences, Beijing 100190, China
2. Beijing Key Laboratory of Mobile Computing and Pervasive Device, Beijing 100190, China
3. School of Computer Science and Technology, University of Chinese Academy of Sciences, Beijing 100049, China
4. National Clinical Research Center for Geriatric Disorders, XuanWu Hospital, Capital Medical University, Beijing 100053, China
 全文: PDF(1191 KB)   HTML
摘要:

针对医疗普适场景下的帕金森病患者高精度用药开-关期检测问题,提出基于腕部姿态的帕金森病开-关期检测方法.利用佩戴在手腕处的运动传感器数据进行姿态解算,得到腕部姿态信息特征,作为卷积神经网络输入进行帕金森病开-关期状态分类. 在医院临床患者测试数据上进行的对比实验表明,与采用运动传感器原始数据的最优结果相比,采用姿态信息能够提升20.3%的检测准确率;与当前最优的网络结构相比,该方法所采用的卷积神经网络在保持相似检测准确率(88.7%)的前提下,将模型参数量降低90.4%. 在医院临床患者自由活动数据上进行的实验表明,该方法能够在非限定动作下预测患者开-关期状态,达到开期91.5%和关期94.4%的准确率.

关键词: 帕金森疾病腕部姿态运动状态评估可穿戴传感器深度学习    
Abstract:

A Parkinson's disease ON/OFF state detection method based on wrist attitude was proposed aiming at the high-precision ON/OFF state detection of Parkinson's disease patients in pervasive healthcare scenarios. The motion sensor data worn on the wrist were utilized to obtain the wrist attitude feature with attitude determination, which was used as the input of the convolutional neural network to classify the ON/OFF state of Parkinson's disease. The comparative experiment on the clinical patient test data shows that using attitude information can obtain a 20.3% accuracy improvement compared with the optimal results of raw sensor data. The convolutional neural network reduced the amount of model parameters by 90.4% while maintaining a detection accuracy of 88.7%. This accuracy is similar to that of the current optimal network structure. Experiments conducted on the free movement data of clinical patients show that the method can predict the patient's ON/OFF state under unrestricted actions, and achieves an accuracy of 91.5% in the ON state and 94.4% in the OFF state.

Key words: Parkinson's disease    wrist attitude    motor state assessment    wearable sensor    deep learning
收稿日期: 2020-09-08 出版日期: 2021-05-07
CLC:  TP 391  
基金资助: 国家自然科学基金资助项目(61972383,61902379);2020年度中国残联课题资助项目(CJFJRRB23-2020);河北省自然科学基金资助项目(F2019207061)
通讯作者: 蒋鑫龙     E-mail: zhangteng19s@ict.ac.cn;jiangxinlong@ict.ac.cn
作者简介: 张腾(1993—),男,硕士生,从事普适计算的研究. orcid.org/0000-0003-1870-1051. E-mail: zhangteng19s@ict.ac.cn
服务  
把本文推荐给朋友
加入引用管理器
E-mail Alert
作者相关文章  
张腾
蒋鑫龙
陈益强
陈前
米涛免
陈彪

引用本文:

张腾,蒋鑫龙,陈益强,陈前,米涛免,陈彪. 基于腕部姿态的帕金森病用药后开-关期检测[J]. 浙江大学学报(工学版), 2021, 55(4): 639-647.

Teng ZHANG,Xin-long JIANG,Yi-qiang CHEN,Qian CHEN,Tao-mian MI,Piu CHAN. Wrist attitude-based Parkinson's disease ON/OFF state assessment after medication. Journal of ZheJiang University (Engineering Science), 2021, 55(4): 639-647.

链接本文:

http://www.zjujournals.com/eng/CN/10.3785/j.issn.1008-973X.2021.04.005        http://www.zjujournals.com/eng/CN/Y2021/V55/I4/639

图 1  典型患者的UPDRS评分与开-关期状态认定
图 2  开-关期检测方法整体流程
图 3  坐标系 $\alpha $至坐标系 $\;\beta$的旋转示意图
图 4  用于开-关期检测的CNN网络结构
图 5  ResNet的残差学习模块
网络 参数量
CNN-4 59 570
AlexNet 1 251 103
ResNet20 623 586
ResNet56 1 723 234
ResNet101 3 097 794
ResNet155 4 747 266
表 1  不同网络参数量
属性 统计值
性别(男∶女) 5∶4
年龄(岁) 62.9±7.6
惯用手(左∶右) 1∶8
帕金森患病时长(年) 3.8±1.9
左旋多巴制剂用药时长(年) 1.5±2.2
开期UPDRS评分 38.2±10.7
关期UPDRS评分 20.3±6.3
表 2  患者人口学统计信息
输入 A Sen Sp AUROC
加速度计 0.68±0.013 0.762±0.030 0.635±0.017 0.771±0.012
MARG 0.555±0.095 0.688±0.451 0.470±0.379 0.586±0.097
姿态信息 0.887±0.008 0.857±0.009 0.890±0.015 0.948±0.008
表 3  腕部姿态信息检测开-关期状态的结果
图 6  各层神经网络的验证集准确率
图 7  传感器原始数据与姿态信息在训练中的性能分析
图 8  患者在整个测试过程中的开-关期预测结果
1 DORSEY E R, CONSTANTINESCU R, THOMPSON J P, et al Projected number of people with Parkinson disease in the most populous nations, 2005 through 2030[J]. Neurology, 2007, 68 (5): 384- 386
doi: 10.1212/01.wnl.0000247740.47667.03
2 ZHANG Z X, ROMAN G C, HONG Z, et al Parkinson's disease in China: prevalence in Beijing, Xian, and Shanghai[J]. Lancet, 2005, 365 (9459): 595- 597
doi: 10.1016/S0140-6736(05)70801-1
3 DAVIE C A A review of Parkinson's disease[J]. British Medical Bulletin, 2008, 86 (1): 109- 127
doi: 10.1093/bmb/ldn013
4 JANKOVIC J Parkinson's disease: clinical features and diagnosis[J]. Journal of Neurology, Neurosurgery and Psychiatry, 2008, 79 (4): 368- 376
doi: 10.1136/jnnp.2007.131045
5 COTZIAS G C, PAPAVASILIOU P S, GELLENE R Modification of Parkinsonism-chronic treatment with L-Dopa[J]. The New England Journal of Medicine, 1969, 280 (7): 337- 345
doi: 10.1056/NEJM196902132800701
6 KEIJSERS N L, HORSTINK M W, GIELEN S C Ambulatory motor assessment in Parkinson's disease[J]. Movement Disorders, 2006, 21 (1): 34- 44
doi: 10.1002/mds.20633
7 HSSAYENI M D, BURACK M A, GHORAANI B. Automatic assessment of medication states of patients with Parkinson's disease using wearable sensors [C]// Proceedings of the 38th Annual International Conference of the IEEE Engineering in Medicine and Biology Society. Orlando: IEEE, 2016: 6082-6085.
8 HSSAYENI M D, BURACK M A, M D, et al. Wearable-based mediation state detection in individuals with Parkinson's disease [EB/OL]. (2019–12–10) [2020–11–02]. https://arxiv.org/abs/1809.06973.
9 SAMà A, PéREZ-LOPEZ C, ROMAGOSA J. Dyskinesia and motor state detection in Parkinson's disease patients with a single movement sensor [C]// Proceedings of the 34th Annual International Conference of the IEEE Engineering in Medicine and Biology Society. San Diego: IEEE, 2012: 1194-1197.
10 HAMMERLA N Y, FISHER J, ANDRAS P, et al. PD disease state assessment in naturalistic environments using deep learning [C]// 29th AAAI Conference on Artificial Intelligence. Austin: AAAI, 2015: 1742-1748.
11 ESKOFIER B M, LEE S I, DANEAULT J F, et al. Recent machine learning advancements in sensor-based mobility analysis: deep learning for Parkinson's disease assessment [C]// Proceedings of the 38th Annual International Conference of the IEEE Engineering in Medicine and Biology Society. Orlando: IEEE, 2016: 655-658.
12 TSIPOURAS M G, TZALLAS A T, FOTIADIS D I, et al. On automated assessment of Levodopa-induced dyskinesia in Parkinson's disease [C]// Proceedings of the 33rd Annual International Conference of the IEEE Engineering in Medicine and Biology Society. Boston: IEEE, 2011: 2679-2682.
13 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
doi: 10.1002/mds.22340
14 O'SUILLEABHAIN P E, DEWEY R B J Validation for tremor quantification of an electromagnetic tracking device[J]. Movement Disorders, 2001, 16 (2): 265- 271
doi: 10.1002/mds.1064
15 BEUTER A, GEOFFROY A, CORDO The measurement of tremor using simple laser systems[J]. Journal of Neuroscience Methods, 1994, 53 (1): 47- 54
doi: 10.1016/0165-0270(94)90143-0
16 ZITO G A, GERBER S M, URWYLER P, et al. Development and pilot testing of a novel electromechanical device to measure wrist rigidity in Parkinson’s disease [C]// Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society. [S. l. ]: IEEE, 2018: 4885–4888.
17 ALLEN D P, PLAYFER J R, ALY N M On the use of low-cost computer peripherals for the assessment of motor dysfunction in Parkinson's disease-quantification of Bradykinesia using target tracking tasks[J]. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2007, 15 (2): 286- 294
doi: 10.1109/TNSRE.2007.897020
18 PASTOR M A, JAHANSHAHI M, ARTIEDA J, et al Performance of repetitive wrist movements in parkinson's disease[J]. Brain, 1992, 115 (3): 875- 891
doi: 10.1093/brain/115.3.875
19 LONINI L, DAI A, SHAWEN N, et al Wearable sensors for Parkinson's disease: which data are worth collecting for training symptom detection models[J]. NPJ Digital Medicine, 2018, 1 (1): 64- 71
doi: 10.1038/s41746-018-0071-z
20 PATEL S, LORINCZ K, HUGHES R, et al Monitoring motor fluctuations in patients with Parkinson's disease using wearable sensors[J]. IEEE Transactions on Information Technology in Biomedicine, 2009, 13 (6): 864- 873
doi: 10.1109/TITB.2009.2033471
21 MANCINI M, PRIEST K C, NUTT J G, et al. Quantifying freezing of gait in Parkinson's disease during the instrumented timed up and go test [C]// Proceedings of the 34th Annual International Conference of the IEEE Engineering in Medicine and Biology Society. San Diego: IEEE, 2012: 1198-1201.
22 VINCE J. Geometric algebra for computer graphics [M]. London: Springer, 2008: 39-48.
23 HEMINGWAY E G, O'REILLY O M Perspectives on Euler angle singularities, gimbal lock, and the orthogonality of applied forces and applied moments[J]. Multibody System Dynamics, 2018, 44 (1): 31- 56
doi: 10.1007/s11044-018-9620-0
24 KRIZHEVSKY A, SUTSKEVER I, HINTON G E ImageNet classification with deep convolutional neural networks[J]. Communications of the ACM, 2017, 60 (6): 84- 90
doi: 10.1145/3065386
25 HE K M, ZHANG X Y, REN S Q, et al. Deep residual learning for image recognition [C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas: IEEE, 2016: 770-778.
[1] 许佳辉,王敬昌,陈岭,吴勇. 基于图神经网络的地表水水质预测模型[J]. 浙江大学学报(工学版), 2021, 55(4): 601-607.
[2] 王虹力,郭斌,刘思聪,刘佳琪,仵允港,於志文. 边端融合的终端情境自适应深度感知模型[J]. 浙江大学学报(工学版), 2021, 55(4): 626-638.
[3] 徐利锋,黄海帆,丁维龙,范玉雷. 基于改进DenseNet的水果小目标检测[J]. 浙江大学学报(工学版), 2021, 55(2): 377-385.
[4] 许豪灿,李基拓,陆国栋. 由LeNet-5从单张着装图像重建三维人体[J]. 浙江大学学报(工学版), 2021, 55(1): 153-161.
[5] 黄毅鹏,胡冀苏,钱旭升,周志勇,赵文露,马麒,沈钧康,戴亚康. SE-Mask-RCNN:多参数MRI前列腺癌分割方法[J]. 浙江大学学报(工学版), 2021, 55(1): 203-212.
[6] 陈巧红,陈翊,李文书,贾宇波. 多尺度SE-Xception服装图像分类[J]. 浙江大学学报(工学版), 2020, 54(9): 1727-1735.
[7] 郑浦,白宏阳,李伟,郭宏伟. 复杂背景下的小目标检测算法[J]. 浙江大学学报(工学版), 2020, 54(9): 1777-1784.
[8] 周登文,田金月,马路遥,孙秀秀. 基于多级特征并联的轻量级图像语义分割[J]. 浙江大学学报(工学版), 2020, 54(8): 1516-1524.
[9] 明涛,王丹,郭继昌,李锵. 基于多尺度通道重校准的乳腺癌病理图像分类[J]. 浙江大学学报(工学版), 2020, 54(7): 1289-1297.
[10] 闫旭,范晓亮,郑传潘,臧彧,王程,程明,陈龙彪. 基于图卷积神经网络的城市交通态势预测算法[J]. 浙江大学学报(工学版), 2020, 54(6): 1147-1155.
[11] 汪周飞,袁伟娜. 基于深度学习的多载波系统信道估计与检测[J]. 浙江大学学报(工学版), 2020, 54(4): 732-738.
[12] 杨冰,莫文博,姚金良. 融合局部特征与深度学习的三维掌纹识别[J]. 浙江大学学报(工学版), 2020, 54(3): 540-545.
[13] 洪炎佳,孟铁豹,黎浩江,刘立志,李立,徐硕瑀,郭圣文. 多模态多维信息融合的鼻咽癌MR图像肿瘤深度分割方法[J]. 浙江大学学报(工学版), 2020, 54(3): 566-573.
[14] 贾子钰,林友芳,张宏钧,王晶. 基于深度卷积神经网络的睡眠分期模型[J]. 浙江大学学报(工学版), 2020, 54(10): 1899-1905.
[15] 王万良,杨小涵,赵燕伟,高楠,吕闯,张兆娟. 采用卷积自编码器网络的图像增强算法[J]. 浙江大学学报(工学版), 2019, 53(9): 1728-1740.