Objects: To evaluate the feasibility and effectiveness of in-bed wearable elbow robot training for motor recovery in patients with early and late subacute stroke. Methods: Eleven in-patient stroke survivors (men/women: 7/4, age: 50.7±10.6 years, post-stroke duration: 2.6±1.9 months received 15 sessions of training over about 4 weeks of hospital stay. During each hourly training, participants received passive stretching and active movement training with motivating games using a wearable elbow rehabilitation robot. Isometric maximum muscle strength (MVC) of elbow flexors and extensors was evaluated using the robot at the beginning and end of each training session. Clinical measures including Fugl-Meyer Upper Extremity motor scale (FMA-UE), Motricity Index for upper extremities scale (MI), Modified Ashworth Scale (MAS) were measured at baseline, after the 4-week training program, and at a one-month follow-up. The muscle strength recovery curve overtime was characterized as a logarithmic learning curve with 3 parameters. Results: At the baseline, participants had moderate to severe upper limb motor impairment (FMA-UE (median [interquartile range]): 28 [18-45]) and mild spasticity in elbow flexors (MAS (median [interquartile range]): 0 [0-1]). After about 4 weeks of training, significant improvements were observed in FMA-UE (p=0.003) and MI (p=0.005), and the improvements were sustained at the follow-up. The elbow flexors MVC significantly increased by 1.93Nm (95%CI: 0.93-2.93Nm, p=0.017), the elbow extensor MVC increased by 0.68Nm (95%CI: -0.05-1.98Nm, p=0.036). Muscle strength recovery curve showed that patients with severe upper limb motor impairment had a greater improvement rate in elbow flexor strength than those with moderate motor impairment. Conclusion: In-bed wearable elbow robotic rehabilitation is feasible and effective in improving biomechanical and clinical outcomes for early and late subacute stroke in-patients. Results from the pilot study suggested that patients with severe motor impairment may benefit more from the robot training compared to those with moderate impairment.

中文翻译：

---

使用可穿戴肘部机器人在早期和晚期亚急性中风中进行床内感觉运动康复：一项初步研究

目的：评估卧床可穿戴肘部机器人训练对早期和晚期亚急性脑卒中患者运动恢复的可行性和有效性。方法：11 名住院卒中幸存者（男性/女性：7/4，年龄：50.7±10.6 岁，卒中后持续时间：2.6±1.9 个月）在住院时间约 4 周内接受了 15 次培训。在每个小时的培训期间, 参与者使用可穿戴的肘部康复机器人接受了被动拉伸和主动运动训练和激励游戏。在每次训练的开始和结束时，使用机器人评估肘部屈肌和伸肌的等长最大肌肉力量 (MVC)。临床措施包括 Fugl -Meyer 上肢运动量表 (FMA-UE)、上肢运动指数 (MI)、改良 Ashworth 量表 (MAS) 在基线测量，在为期 4 周的培训计划之后，并在 1 个月的后续行动中。肌肉力量恢复曲线超时被表征为具有3个参数的对数学习曲线。结果：在基线时，参与者有中度至重度上肢运动障碍（FMA-UE（中位数 [四分位距]）：28 [18-45]）和肘屈肌轻度痉挛（MAS（中位数 [四分位距]）： 0 [0-1]）。经过大约 4 周的训练，观察到 FMA-UE (p=0.003) 和 MI (p=0.005) 有显着改善，并且在随访中持续改善。肘屈肌 MVC 显着增加了 1.93Nm（95%CI：0.93-2.93Nm，p=0.017），肘伸肌 MVC 增加了 0.68Nm（95%CI：-0.05-1.98Nm，p=0.036）。肌力恢复曲线显示，重度上肢运动障碍患者的肘屈肌力量改善率高于中度运动障碍患者。结论：卧床可穿戴肘部机器人康复在改善早期和晚期亚急性脑卒中住院患者的生物力学和临床结果方面是可行和有效的。试点研究的结果表明，与中度障碍患者相比，严重运动障碍患者可能从机器人训练中受益更多。