摘要：

目的 采用3D动作捕捉系统对帕金森病患者上肢运动迟缓进行数字化测评，以弥补人工测评不足。方法 选择2019年6-12月临床确诊的19例帕金森病患者和10例同期健康体检者（对照组）为受试对象，在安装有3D动作捕捉系统的房间内完成统一帕金森病评价量表第三部分（UPDRS-Ⅲ）测评，受试者拇指和食指固定反光标记物，以同时完成对指这一动作的数字化测评，从中提取最大振幅（amp）、间隔时间（intval）、加速度过零次数（zc）、正向最大速度（v1）和负向最大速度（v2）共5项参数；绘制受试者工作特征（ROC）曲线评价intval、zc单项或联合测评对帕金森病上肢运动迟缓及对指评分1分时的诊断价值。结果 帕金森病组UPDRS-Ⅲ评分（33.79 ±13.14）分，运动迟缓评分（13.61 ±5.26）分，上肢运动迟缓评分（6.61 ±2.65）分，对指评分（1.68 ±0.78）分。帕金森病组intval（P=0.001）和zc（P=0.004）高于正常对照组。intval和zc单项对帕金森病上肢运动迟缓的ROC曲线显示，曲线下面积分别为0.858（95% CI：0.743~0.973，P=0.000）和0.799（95% CI：0.642~0.955，P=0.008），灵敏度为0.737、0.921，特异度为0.900、0.600；intval与zc联合评价，曲线下面积为0.858（95% CI：0.752~0.964，P=0.001），灵敏度0.711、特异度0.100。结论 利用3D动作捕捉系统能够实现帕金森病上肢运动迟缓的数字化测评，在一定程度上弥补人工测评的缺陷；从中提取的运动参数intval和zc具有较高的敏感性和特异性，有望成为帕金森病早期筛查的特征性指标。

关键词: 帕金森病, 运动障碍, 上肢, ROC曲线

Abstract:

Objective The 3D motion capture system was used to measure the upper limb bradykinesia in Parkinson's disease (PD) to make up for the lack of artificial measurement. Methods Ninteen patients with PD (PD group) and 10 healthy subjects (control group) were recruited from June to December 2019 in Tianjin Huanhu Hospital. All the subjects completed the Unified Parkinson's Disease Rating ScaleⅢ (UPDRS-Ⅲ) evaluation in the room equipped with 3D motion capture system. The thumb and index finger of the subjects were fixed with reflective markers to collect motion information. In this study, only the action of finger tapping was evaluated digitally, and five parameters were extracted from the evaluation results:1) average maximum amplitude of finger taps (amp). 2) Average finger tapping interval (intval). 3) Average zero-crossing occurrences of acceleration (zc). 4) Average of the forward maximum velocity (v1). 5) Average of the negative maximum velocity (v2). Using receiver operating characteristic curve (ROC) to evaluate the value of intval, zc and their combination indexes in the diagnosis of upper limb bradykinesia in PD. Results In PD group, UPDRS-Ⅲ score was 33.79 ±13.14, bradykinesia score was 13.61 ±5.26, upper limb bradykinesia score was 6.61 ±2.65, finger tapping score was 1.68 ±0.78. Intval (P=0.001) and zc (P=0.004) in PD group were higher than those in control group, and the difference was statistically significant. The ROC curves of intval and zc for upper limb bradykinesia showed that the AUC of them were 0.858 (95% CI:0.743-0.973, P=0.000) and 0.799 (95% CI:0.642-0.955, P=0.008); the sensitivity was 0.737, 0.921, and the specificity was 0.900 and 0.600, respectively. After the combination of intval and zc, the AUC was 0.858 (95% CI:0.752-0.964, P=0.001), the sensitivity and specificity were 0.711 and 0.100, respectively. Conclusions The digital evaluation of upper limb bradykinesia in PD can be realized by using 3D motion capture system, which makes up for the deficiency of manual evaluation to a certain extent. The extracted parameters intval and zc have high sensitivity and specificity, which are expected to be used as characteristic indicators for early screening of PD.

Key words: Parkinson disease, Motor disorders, Upper extremity, ROC curve