2026 年第 1 期 第 21 卷

任务导向型上肢康复机器人对脑卒中后共同运动期偏瘫患者上肢运动功能恢复的最佳处方探索分析

Exploration and analysis of the optimal prescription of task-oriented upper limb rehabilitation robot for upper limb motor function recovery in hemiplegic patients at the synergistic movement stage after stroke

作者：程迎胡子木任国荣高晓平

英文作者：Cheng Ying Hu Zimu Ren Guorong  Gao Xiaoping

单位：安徽医科大学第一附属医院康复医学科，合肥230022

英文单位：Department of Rehabilitation Medicine the First Affiliated Hospital of Anhui Medical University Hefei 230022 China

关键词：共同运动期偏瘫；脑卒中；任务导向型上肢康复机器人；最佳处方

英文关键词：Hemiplegiaatsynergisticmovementstage;Stroke;Task-orientedupperlimbrehabilitationrobot;Optimalprescription

• 摘要：

• 目的 探究任务导向型上肢康复机器人对脑卒中后共同运动期偏瘫患者上肢运动功能恢复的最佳处方。方法 选取2024年4月1日至2025年4月30日在安徽医科大学第一附属医院康复医学科住院治疗，处于共同运动期的112例脑卒中后上肢偏瘫患者。通过随机数字表法将患者分为对照组、观察A组、观察B组和观察C组，每组28例。对照组在基础药物治疗基础上进行常规康复训练，观察A、B、C组在基础药物治疗基础上进行任务导向型上肢康复机器人低强度、中等强度和高强度的训练，干预周期4周。对比各组患者在干预前（T0）、干预2周（T1）、干预4周（T2）的上肢痉挛程度［改良Ashworth痉挛量表（MAS）评分］、上肢运动功能［Fugl-Meyer上肢运动功能评定量表（FMA-UE）评分］、关节活动范围以及轨迹跟踪误差，并采用线性回归分析及限制性立方样条（RCS）模型开展康复运动训练强度与上肢功能改善的量效分析。结果 整体分析发现，各组MAS评分、FMA-UE评分、屈肌共激活率、Brunnstrom分期、运动空间范围及轨迹跟踪误差随时间均有显著改善，组间差异及交互效应差异均有统计学意义（均P<0.05）。进一步比较发现，观察A、B、C组随时间延展MAS评分、屈肌共激活率［T0、T1、T2观察A组分别为（79±4）%、（74±6）%、（67±5）%，观察B组分别为（79±7）%、（68±7）%、（61±4）%，观察C组分别为（81±6）%、（77±4）%、（69±6）%］逐步下降，FMA-UE评分、运动空间范围逐步上升，轨迹跟踪误差逐步下降（均P<0.05）；对照组T2较T0亦有一定改善（均P<0.05），但幅度不及观察组；T1、T2时点观察组均优于对照组（均P<0.05），观察B组改善最显著，优于A组和C组（均P<0.05），A组与C组差异均无统计学意义（均P>0.05）。Brunnstrom分期在B组T2时点提升最明显。RCS分析显示干预强度与FMA-UE评分呈倒U型关系，中等强度组达到最优干预效果。结论 基于任务导向型上肢康复机器人开展康复训练可显著改善脑卒中后共同运动期偏瘫患者的上肢痉挛程度和运动功能，降低屈肌共激活率，提高关节活动范围，减少轨迹跟踪误差，且最佳康复运动训练处方为单次训练时间30 min，5次/周。

• Objective To explore the optimal prescription of task-oriented upper limb rehabilitation robot for upper limb motor function recovery in hemiplegic patients at the synergistic movement stage after stroke. Methods A total of 112 patients with upper limb hemiplegia after stroke at the synergistic movement stage, who were hospitalized in the Department of Rehabilitation Medicine, the First Affiliated Hospital of Anhui Medical University from April 1, 2024 to April 30, 2025, were selected. The patients were divided into control group, observation group A, observation group B, and observation group C by random number table method, with 28 cases in each group. The control group received routine rehabilitation training on the basis of basic drug treatment, while observation groups A, B, and C received low-intensity, moderate-intensity, and high-intensity task-oriented upper limb rehabilitation robot training respectively on the basis of basic drug treatment, with an intervention cycle of 4 weeks. The upper limb spasticity ［modified Ashworth scale (MAS) score］, upper limb motor function ［Fugl-Meyer assessment of upper extremity (FMA-UE) score］, range of motion, and trajectory tracking error were compared among the groups before intervention (T0), at 2 weeks of intervention (T1), and at 4 weeks of intervention (T2). Linear regression analysis and restricted cubic spline (RCS) model were used to conduct a dose-response analysis between the intensity of rehabilitation exercise training and the improvement of upper limb function. Results Overall analysis showed that the MAS score, FMA-UE score, flexor co-activation rate, Brunnstrom stage, motor space range, and trajectory tracking error in each group improved significantly over time, with statistically significant differences between groups and interaction effects (all P<0.05). Further comparisons revealed that in observation groups A, B, and C, the MAS score and flexor co-activation rate ［observation group A at T0, T1, and T2: (79±4)%, (74±6)%, (67±5)% respectively; observation group B: (79±7)%, (68±7)%, (61±4)% respectively; observation group C: (81±6)%, (77±4)%, (69±6)% respectively］ decreased gradually, while the FMA-UE score and motor space range increased progressively, and the trajectory error decreased stepwise (all P<0.05). The control group also showed certain improvements at T2 compared with T0 (all P<0.05), but the magnitude was less than that of the observation groups. At T1 and T2, all observation groups outperformed the control group (all P<0.05), with observation group B showing the most significant improvements, which were superior to those of groups A and C (all P<0.05), while no statistically significant differences were found between group A and group C (all P>0.05). The Brunnstrom stage showed the most significant improvement in group B at T2. RCS analysis revealed an inverted U-shaped relationship between intervention intensity and FMA-UE score, with the moderate-intensity group achieving the optimal intervention effect. Conclusion  Rehabilitation training based on task-oriented upper limb rehabilitation robot can significantly improve upper limb spasticity and motor function, reduce flexor co-activation rate, increase range of motion, and decrease trajectory tracking error in hemiplegic patients at the synergistic movement stage after stroke. The optimal rehabilitation training prescription is 30 minutes per session, 5 times a week.