中文摘要
目前中風病人在接受復健治療，都是由醫師或是物理治療師以徒手或輔具施予各種不同的治療手法以誘發原有的自主能力，這些手法包括引導病患做出動作，或是在病患作動的過程中給予適當阻力或助力，最後再以定性量測或觀察作為療效的評估。本研究目的是要運用自製的機器人於神經復健，由機器人來達成物理治療師的手法，並透過在機器人上的感測器量測過程中的相關物理訊號，由準確的量化數據作為療效的評估。
本研究設計的機器人是在平面上運動，能應用於病患肘關節的屈曲與伸展運動，在整個作動過程中，透過模糊控制器達成準確的位置與力量控制。而在臨床實驗中，分別選擇5位正常人與5位偏癱的中風病人做為受測者，讓其沿著規劃的圓形軌跡做自主運動，並在其運動方向給予適當的阻力，由所量測到的肱二頭肌與肱三頭肌的肌電訊號，來評估其運動的協調性，另外於其作動過程中施予一外在的干擾負荷，來估算其手臂肢體剛性，以此做為運動控制能力的量化指標。
本研究以作動過程中所量測到的相關生理訊號，訂定肌電訊號指標與關節剛性指標來評估病人的復健成效。在肌電訊號指標方面，由統計分析可以驗證出病人健側的運動協調性優於患側，但是無法應用於不同病人之間的比較。在剛性指標方面，由估測出的肘關節與肩關節剛性參數，可判斷常人與病人之間運動控制能力的差異，且由人體實驗驗證此指標與臨床上用的布朗斯壯指標有高度相關性，甚至能較其對病人的動作表現做出更精細的分類。總之，本研究驗證了機器人在復健應用上的可行性，並得到客觀的量化指標來評估病人復健的成效。

Abstract
Physical therapists use various facilitation patterns by arms or instruments for the rehabilitation of stroke patients. The movement patterns include guiding movements of patient’s limbs and applying resistant or assistant force when active movements are performed. By qualitative assessments and observations, physical therapists can evaluate progress of rehabilitation. The goal of this thesis is to develop a robot for rehabilitation of patients with neuro-muscular disorders by performing various facilitation patterns. The robot equipped with a six-axis load cell can measure biomechanical variables such as joint displacements and reactive forces exerted by patient’s limbs. These quantitative data can be used for the assessment of rehabilitation.
The robot is designed for two-dimensional motion in a planar workspace. It is designed for application in the flexion and extension of elbow joints. During testing process, the robot uses a fuzzy logic controller to realize the position and force control. In the clinical experiment, the system has been used to test five normal subjects and five stroke patients. The subjects can move voluntarily along the circle path and the robot applies appropriate resistant force against the moving direction. The system can record the EMG signals of biceps brachii and triceps brachii in the circle exercise. A disturbance force is applied during the circle exercise and the stiffness of the upper arm motor control system is estimated.
Two quantitative biomechanical indices are defined. One is the ratio between agonist and antagonist EMG and the other is the shoulder and elbow stiffness sensed by the robot. For the EMG index, it can prove statistically that the patient’s intact side is superior to the affected side in the coordination of motion. But it can not be used to contrast the effect among different patients. For joint stiffness index, it can evaluate the difference of the capability of motion control between normal subjects and stroke patients. A high correlation coefficient between the stiffness index and the Brunnstrom’s stage index is found from the experimental data. The stiffness index may have better classification capability for the assessment of motor control ability. From the results, one might conclude that the robot is applicable for rehabilitation of stroke patients and quantitative indices are feasible for assessing the progress of neural rehabilitation in stroke patients.

參考文獻
1. K. A. Sawner and J. M. LaVigne, Brunnstrom’s Movement Therapy in Hemiplegia : a Neurophysiological Approach, J. B. Lippincott, USA, 1992.
2. M. L. Latash, Control of Human Movement, Human Kinetics Publishers, USA, 1993.
3. R. F. Kirsch and A. M. Acosta, “Techniques for Characterizing Elbow and Shoulder Mechanics and Neural Control in Normal and Disabled Subjects,” IEEE-EMBC and CMBEC Theme 5: Neuromuscular Systems/Biomechanics, pp. 1291-1292, 1995.
4. T. Noritsugu and T. Tanaka, “Application of Rubber Artificial Muscle Manipulator as a Rehabilitation Robot,” IEEE/ASME Tran. on Mechatronics, vol. 2, no. 4, Dec. 1997.
5. H. I. Krebs, N. Hogan, M. L. Aisen and B. T. Volpe, “Robot-Aided Neurorehabilitation,” IEEE Tran. on Rehabilitation Engineering, vol. 6, pp. 75-87, Mar. 1998.
6. C. J. Feng, A. F. T. Mak and C. S.J. Xiao, “Electromyographic Analysis of Voluntary Elbow Movements in Subjects with Spasticity,” IEEE-EMBC and CMBEC Theme 5: Neuromuscular Systems/Biomechanics, pp. 1295-1296, 1997.
7. J. A. Cozens, “Robotic Assistance of an Active Upper Limb Exercise in Neuological Impaired Patients,” IEEE Tran. on Rehabilitation Engineering, vol. 7, pp. 254-256, June 1999.
8. 林棟煌，肘關節神經復健用機器人之改良和臨床測試，國立成功大學機械工程系碩士論文，2001。
9. H. Asada and K. Youcef-Toumi, “Analysis and Design of a Direct-Drive Arm with a Five- Bar-Link Parallel Drive Mechanism,” Tran. ASME Journal of Dynamic System, Measurement, and Control, vol. 106, pp. 255-230, Sep. 1984.
10. B. Tyldesley and J. I. Grieve, Muscles, Nerves and Movement, Blackwell Science, USA, 1996.
11. N. Palastanga, D. Field and R. Soames, Anatomy and Human Movement, Butterworth-Heineman, UK, 1994.
12. 林裕人，多頻道表面肌電訊號之高頻頻譜分析，國立成功大學醫學工程系碩士論文，1994。
13. C. Y. Yeh, “Evaluate the Effects of Prolonged Muscle Stretch on Motor neuron Excitability with Spastic Hemiplegia,” Master Thesis, Institute of Medicine, Chung Shan Medical and Dental College, June 1994.
14. T. Moritani, H. A. Devries, “Reexamination of the Relationship Between the Surface Integrated Electromyogram and Force of Isometric Contraction,” American J. of Physical Medicine, vol. 57, no. 6, pp. 263-277,1978.
15. R. F. Kirsch, R. E. Kearney and P. E. Crago, “Dynamic Stiffness and Measurement and the Control of Movement,” IEEE, pp. 428-429, 1994.
16. E. J. Perreault, R. F. Kirsch, Patrick E. Crago, “Effects of Shoulder Stability on Endpoint Stiffness,” IEEE Eng. in Medicine and Biology, pp. 53-57, Dec. 2000.
17. H. Asada and J.-J. E. Slotine, Robot Analysis and Control, A Wiley Interscience Publication, Canada, 1986.