近期研究 Recent Research​

Muscle Activation Based Active-assistive Control for Upper Limb Rehabilitation Robot

        It is necessary to train patients to relearn the correct muscle synergy during the rehabilitation process. This research proposes a muscle activation-based active-assistive control strategy for upper limb exoskeleton rehabilitation robots. We monitor the surface electromyography signals of the subject's primary muscles associated with the target joints under the specific rehabilitation task, convert the recorded signal to the activation level and evaluate the normality of the subject's muscle activation distribution. The results show that the muscle activation level of the user is accurately detected, and the user can obtain more assistance through correct muscle activation during the rehabilitation process.

Learning-based Home Rehabilitation Assessment System by Inertial Measurement Units Trajectory ReconstructionLearning-based Home Rehabilitation Assessment System by Inertial Measurement Units Trajectory Reconstruction

        Human motion capture and reconstruction is a topic that has gained a lot of attention in recent years as it has a wide range of applications. This study focuses on upper limb motion capture and trajectory reconstruction, which uses inertial measurement units to recover rehabilitation exercise. In addition, by taking advantage of big data, we use machine learning to learn the scoring mechanism of the therapist and give feedback to the patient.

Muscle Fatigue Detection System Based on sEMG Assisted with EEG

        During rehabilitation training, patients often ignore muscle fatigue and continue to do high-intensity training while pursuing rapid recovery. This situation may weaken the effect of rehabilitation treatment and even endanger the personal safety of patients. We propose a fatigue monitoring system based on surface electromyography signals for muscle isotonic contraction. At the same time, a non-invasive electroencephalogram will be used to monitor the state of sustained attention to reduce the interference of continuous attention state on muscle fatigue detection and increase the accuracy of the monitoring system.

IMU-based Estimation of Lower Limb Motion Trajectory with Graph Convolution Network

        In the field of rehabilitation, motion capture systems are often used to collect the motion information of the patient when he/she is performing rehabilitation tasks. However, the sensor measurements will be affected by signal noise, bias and inaccurate gyroscope estimation during long-term wearing, which may cause drift problems. In this research, we propose a deep learning architecture that uses effective information collected by inertial sensors to predict the relative position of each joint of human’s lower limbs in 3D space of the future motion sequence. The motion prediction result from the deep learning model is used to alleviate the tracking error caused by reliance only on inertial sensor.

Active-Assistive Control System with Slacking Behavior Prevention for Upper Limb Rehabilitation Exoskeleton Robot

        In this research, a participation-based active-assistive control strategy and a task difficulty adjustment algorithm are proposed for upper-limb rehabilitation exoskeleton robot, NTUH-II. we propose a subject's participation monitoring mechanism to translate the surface electromyography into the active participation level in real-time. The proposed participation-based aggregation method has the ability to provide assistance according to the active intention of the patient, assistive motion, and participation level to achieve slacking behavior prevention. The experiment result show that the proposed method can prevent the subject's slack behavior while assisting the subject in achieving the rehabilitation goal accurately. Additionally, the effectiveness of the task adjustment algorithm is also shown in the experiment results.

Velocity Field based Active-Assistive Control for Upper Limb Rehabilitation Exoskeleton Robot

        Active-assistive control in related works generally limits subjects to perform rehabilitation tasks with particular time instants. In this paper, we propose a velocity field based active-assistive control system to solve this issue. To obtain subjects' active intention of motion, a Kalman filter based interactive torque observer is utilized. A position-dependent velocity field which can be automatically generated via the task motion pattern is designed to provide time-independent assistance. The proposed integration method combines the active and assistive motion based on the performance and the involvement of subjects to make them perform the task actively and accurately. The stability of the proposed system is verified by Lyapunov analysis. The experiment results show that both the execution time and the subjects' exertion are reduced when performing given tasks compared with the related work. In addition, the proposed system can retain subjects' active intention while assisting them to accomplish the task accurately.

Automatic Compensatory Movement Detection System for Upper Limb Robot Rehabilitation

     In this research, we develop a skeleton-based indoor rehabilitation posture monitoring system which is used to automatically detect whether the patient’s compensatory movement occurs during the rehabilitation exercise, and to give real-time visual and auditory feedback to prompt the patient to correct the abnormal posture. The system and model proposed in this research are established on the upper limb exoskeleton rehabilitation robot NTUH-II. The results show that the classification accuracy of the proposed method is better than the relevant studies. Besides, through the design of the window voting method, the real-time performance of compensatory movement detection is well, so that the system can be applied to the rehabilitation therapies.

Deep Learning based Motion Prediction for Exoskeleton Robot Control in Upper Limb Rehabilitation

         In this study, guide control of our recently developed 8 degrees-of-freedom (DOFs) upper limb rehabilitation exoskeleton, named NTUH-II, is proposed. The human arm dynamics and muscle activity can be measured by two Myo sensors. With such setting, the motion information of the user can be predicted by deep learning model from the measured arm dynamics and EMG data and be used as the desired motion trajectory of the exoskeleton. As a result, robot arm will follow the movement from the either side of user’s arm performed in reality. Various experiments have been conducted on three healthy subjects to verify the performance of the proposed guide control, and the results show that the proposed control scheme can reduce the mean absolute error and delay time of robot arm in guide mode.

Sensorless Exoskeleton Robot Control with Friction Estimation for Upper Limb Rehabilitation

        In this study, a sensorless active control of upper limb rehabilitation exoskeleton, named NTUH-II is proposed. Due to mechanical structure of NTUH-II, the friction behavior is more complicated than the commercial robot. However, the accurate friction model is crucial for constructing an interaction torque observer using Kalman filter, which is used to acquire human intention. Furthermore, the human intention needs to be take into consideration to derive the desired motion trajectory of the exoskeleton. The result shows that it can improve the smoothness of the motion and reduce the subject’s effort compared with the related works. Moreover, compared with the use of electromyography sensor and force/torque sensor, this method can get rid of additional sensors while achieving good performance.

Interactive Torque Observer based Exoskeleton
​Robot Control for Upper Limb Rehabilitation

        In this research, we construct the interactive torque observer based on robot dynamic model and measurements of encoder readings and motor torques. Then, based on the dynamic model and interactive torque, we propose a novel interactive torque observer based control for exoskeleton rehabilitation robot for realizing passive, active-assistive, active mode control. Several experiments have been conducted on three subjects which verify the performance of the proposed interactive torque observer based controller. The results show that the proposed control method can manipulate steadily in passive and active-assistive mode exercises. Moreover, the performance in active mode exercises show that it can improve the smoothness and reduce the subject’s effort comparing with one of the related work.

Task-oriented Upper Limb Rehabilitation through
EMG Sensing-based Intention

        We proposed a new robot-assisted task-oriented rehabilitation therapy with electromyography (EMG) sensing-based motion intention recognition model (MIRM) for upper limb has been developed and implemented on NTUH-II. With the concept of motor learning theory, clinical study has shown that task-oriented rehabilitation is an efficient way to relearn the activities of daily living (ADL). The proposed MIRM can predict the intended motion direction of human in a task from the electrical activation of muscles. By the proposed MIRM along with our previous proposed interactive torque observer based active controller, not only passive mode therapy but also active and active-assistive mode task-oriented rehabilitation therapies are available for patients. Various experiments have been conducted on three healthy subjects to verify the performance of the proposed method, and the results show that the performance of the proposed method exceeds the state-of-the-art works and improves the performance of the active control.

研究概況  Introduction

        本組致力於開發醫療復健及照護用之輔具及其相關軟硬體系統設計與應用，所開發之上肢復健機器人可配合治療師進行個人化的復健療程規劃與復健動作的執行。原理藉由帶動病患的肩、肘等上肢關節，偵測於復健過程中病患的生理訊號，並搭配人機互動遊戲介面，讓病患在訓練過程中可以獲得鼓勵、增加信心，也可以藉由復健機器人即時的生理資訊擷取及互動遊戲所獲得的分數加以量化分析治療之成效，有利於治療師長期觀察病患恢復的情形及日後追蹤診斷的依據，並可搭配病患復原的情況即時調整治療的方式。本研究團隊致力於系統設計、控制方法設計、運動軌跡規劃、人機互動、在治療過程中的安全考量。
 
        Led by Professor Li-Chen Fu, the Rehabilitation Robotics Group in the Department of Electrical Engineering at the National Taiwan University has been working on biomedical engineering projects related to physical rehabilitation and the development of medical instruments since 2005. The main focus is upper-limb exoskeleton type rehabilitation robot and its related technology of human-robot interaction, which were retrain motor movement of the limb. The aim of our team is to design both the hardware and software system to help stroke patients achieve as much functional independence as possible and to maintain quality of life by robot-aided assessment and training. The principle of the upper limb rehabilitation robot is to drive the patient's shoulder, elbow and other upper limb joints. Moreover, combining with the interactive game is tried to encourage patient to participant more and also increase patient confidence. In addition, real-time physiological information and game score can be analyzed and qualified as the effeteness of the treatment. With the robot-aided assessment of every time treatment, the therapist can clearly know the patients recover situation and adjust the proper therapies. We were interested in the system design, control strategies, virtual reality technologies, and safety issue during development. The trajectory desired and the kinematic of human-robot interaction were also considered.
 
有鑑於病患受傷程度不同，本組研究團隊開發出不同的復健療程：

  1. 被動療程主要是針對上肢無任何肌力或者關節沾黏的病人，此療程中，上肢復健機器人將會全程帶動病患的上肢做伸展的動作(例如:肩關節屈曲、肩關節外展)，使病人可以利用重複性的訓練建立起神經與肌肉的連結以及關節的拉筋。
  2. 輔助療程主要是針對上肢無足夠肌力的病人，此療程中，上肢復健機器人會有一個預先規劃的運動軌跡，病人需帶動上肢復健機器人沿著此軌跡運動，當病人有落後於軌跡(肌無力等)的情況下，上肢復健機器人則會給予病人輔助力使運動軌跡追回預先規劃好的軌跡，目的是增強病人的肌力。
  3. 主動療程主要是針對已有一定肌力但控制肌肉能力較差的病人，此療程中，我們利用肌電訊號感測器與本研究團隊開發的交互作用扭矩觀測器去判斷病人的運動意圖，藉由此“意圖”去控制上肢復健機器人，目的則是為了訓練病人對肌肉的控制程度，且搭配遊戲使治療過程有別於傳統物理治療的乏味，增添許多趣味性。
  4. 提出一新的環形療程，概念來自於傳統復健療程中在水平面畫圓的運動，本研究團隊加以改良並設計出在垂直面進行畫圓的療程，目的則是為了使肩關節可以雙軸同動，增加肌肉的協調性。​

​In view of the different degree of injury to the patient, our research team developed different rehabilitation treatments :
  1. Passive therapy is mainly aimed at the patients who suffer from upper limb without any muscle strength or joint adhesion. During this therapy, the upper limb rehabilitation robot will push the patient's upper limbs doing stretch exercise (ex: shoulder flexion, shoulder joint abduction). So that, patients can use repetitive training to establish nerve and muscle connection and joint tension.
  2. Assistant therapy mainly aim at the patient whose upper limb has not enough muscle strength. In this therapy, the upper limb rehabilitation robot will have a predefined trajectory, patients need to drive the upper limb rehabilitation robot along this trajectory. When the patient lag behind the trajectory (muscle weakness, etc.), the upper limb rehabilitation robot will give the patient assistance force to recover the trajectory. The therapy is designed to enhance the patient's muscular strength.
  3. Active therapy mainly aims at patients who already have some muscle strength but have poor control of muscle ability. During this therapy, we use EMG sensors and the interactive torque observer developed by our research team to judge (predict) the patient's motion intention. We use this signal to control the upper limb rehabilitation robot in order to train the patient’s control ability on muscle, and cooperating with the game can make the treatment process more interesting than the traditional physical treatment.
  4. We proposed a new circumduction therapy, the concept come from the traditional rehabilitation treatment which is circumduction exercise in the horizontal plane. Our team improves and designs circumduction exercise in the vertical plane. The purpose of this therapy is to make the shoulder joint can be movie in two axes increasing the coordination of muscle.

        目前本研究成果已取得中華民國發明專利並通過臺大醫院醫學倫理委員會臨床試驗及衛生署臨床試驗許可，本系統現於臺大醫院復健部進行中風患者之相關臨床試驗。

       This research is cooperating with National Taiwan University Hospital and is sponsored by National Science Council, R. O. C. and by National Taiwan University Hospital. The clinical trial was approved by Department of Health (DOH), R. O. C. and Research Ethics Committee B of NTUH since 2009.

 Journal

  1. S.-H. Hsu and L.-C. Fu, “Adaptive decentralized control of robot manipulators driven by current-fed induction motors,” IEEE/ASME Transactions on Mechatronics, Vol. 10, No. 4, pp. 465-468, 2005.
  2. F.-Y. Hsu, and L.-C. Fu, " Intelligent robot deburring using adaptive fuzzy hybrid position/force control," IEEE Trans. Robot. Autom., Vol.16, No.4, pp.325-335, 2000.
  3. K.-Y. Lain, L.-S. Wand, and L.-C. Fu, " A skew-symmetric property of rigid-body systems," Syst. Control Lett., No. 33, pp. 187-197, 1998.
  4. K.-Y. Lian, L.-S. Wang, and L.-C. Fu, " Globally valid adaptive controller of mechanical system ," IEEE Trans. Automat. Contr., Vol. 42, No. 8, pp. 1149-1154, 1997.
  5. Yang, J.-H., F.-L. Lian, and L.-C. Fu, "Nonlinear adaptive control of flexible-link manipulators," IEEE Trans. on Robot. Autom., Vol. 13, No. 1, pp. 140-148, 1997.
  6. J.-H. Jean, and L.-C. Fu, "An adaptive control scheme for coordinated multi-manipulator systems," IEEE Trans. Robot. Autom., Vol. 9, No.2, pp. 226-231, (SCI) 1993.
  7. J.-H. Jean, and L.-C. Fu, "Adaptive hybrid control strategy for constrained robots," IEEE Trans. Automat. Contr., Vol. 38, No. 4, pp. 598-603, 1993.
  8. K.-Y. Lian, J.-H. Jean, and L.-C. Fu, "Adaptive force control for mechanical systems with flexible joints: A single link case," IEEE Trans. Robot. Autom., Vol. 7, No. 4, pp. 540-545, (SCI) 1991.
  9. T.-L. Liao, L.-C. Fu, and C.-F. Hsu, "Adaptive robust tracking of nonlinear systems and with an application to a robotic manipulator," Syst. Control Lett., No. 15, pp. 339-348, (SCI) 1990.
  10. L.-C Fu, and T.-L. Liao, "Globally stable robust tracking of nonlinear systems using variable structure control and with an application to a robotic manipulator," IEEE Trans. Automat. Contr., Vol. 35, No. 12, pp. 1345-1350, (SCI) 1990.

Conference​

  1. E.-Y. Chin, Y.-L. Chen, T.-C. Chien, M.-L Chiang, L.-C. Fu, J.-S. Lai and L. Lu, ''Velocity Field based Active-Assistive Control for Upper Limb Rehabilitation Exoskeleton Robot,'' IEEE International Conference on Robotics and Automation (ICRA), 2020.
  2. H.-Y. Li, L.-Y. Chien, H.-Y. Hong, S.-H. Pan, C.-L. Chiao, H.-W. Chen, L.-C. Fu, and J.-S. Lai, “Active control with force sensor and shoulder circumduction implemented on exoskeleton robot NTUH-II,” IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2016.
  3. C.-H. Lin, W.-M. Lien, W.-W. Wang, S.-H. Chen, C.-H. Lo ,S.-Y. Lin , L.-C. Fu ,J.-S. Lai , “NTUH-II Robot Arm with Dynamic Torque Gain Adjustment Method for Frozen Shoulder Rehabilitation,” IEEE/RSJ International Conference on Intelligent Robots and Systems, 2014
  4. G. D. Lee, W. W. Wang, K. W. Lee, S. Y. Lin, L. C. Fu, and J. S. Lai, “Arm exoskeleton rehabilitation robot with assistive system for patient after stroke,” International Conference on Control, Automation and Systems Proceedings (ICCAS), 2012.
  5. L. C. Hsu, W. W. Wang, G. D. Lee, Y. W. Liao, L. C. Fu, and J. S. Lai, “A gravity compensation-based upper limb rehabilitation robot,” American Control Conference Proceedings (ACC), 2012.
  6. W. W. Wang and L. C. Fu, “Mirror therapy with an exoskeleton upper-limb robot based on IMU measurement system,” IEEE International Workshop on Medical Measurements and Applications Proceedings (MeMeA), pp. 370-375, 2011.
  7. B. C. Tsai, W. W. Wang, L. C. Hsu, L. C. Fu, and J. S. Lai, “An articulated rehabilitation robot for upper limb physiotherapy and training,” IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp. 1470-1475, 2010.

Master Thesis​

  1. Chia-Chun Huang, "Muscle Activation Based Active-assistive Control for Upper Limb Rehabilitation Robot," Master thesis, Dept. Elect. Eng., National Taiwan Univ., Taiwan, 2021
  2. I-Ju Yang, "Learning-based Home Rehabilitation Assessment System by Inertial Measurement Units Trajectory Reconstruction," Master thesis, Dept. Elect. Eng., National Taiwan Univ., Taiwan, 2021.
  3. Xiao-He Wang, "Muscle Fatigue Detection System Based on sEMG Assisted with EEG," Master thesis, Dept. Elect. Eng., National Taiwan Univ., Taiwan, 2021.
  4. Tzu-Chieh Chien, "Active-Assistive Control System with Slacking Behavior Prevention for Upper Limb Rehabilitation Exoskeleton Robot," Master thesis, Dept. Elect. Eng., National Taiwan Univ., Taiwan, 2020.
  5. En-Yu Chia, "Velocity Field based Active-Assistive Control for Upper Limb Rehabilitation Exoskeleton Robot," Master thesis, Dept. Elect. Eng., National Taiwan Univ., Taiwan, 2019.
  6. Wei-Hsuan Chen, "Automatic Compensatory Movement Detection System for Upper Limb Robot Rehabilitation," Master thesis, Dept. Elect. Eng., National Taiwan Univ., Taiwan, 2018.
  7. Jialiang Ren, "Deep Learning based Motion Prediction for Exoskeleton Robot Control in Upper Limb Rehabilitation," Master thesis, Dept. Elect. Eng., National Taiwan Univ., Taiwan, 2018.
  8. Lee-Kai Liu, "Sensorless Exoskeleton Robot Control with Friction Estimation for Upper Limb Rehabilitation,"  Master thesis, Dept. Elect. Eng., National Taiwan Univ., Taiwan, 2018.
  9. Hao-Ying Li, ”Intention based Active Exoskeleton Control for Upper Limb Rehabilitation Therapy,” Master thesis, Dept. Elect. Eng., National Taiwan Univ., Taiwan, 2017.
  10. Li-Yu Chien, ”Interactive Torque Observer based Exoskeleton Robot Control for Upper Limb Rehabilitation,” Master thesis, Dept. Elect. Eng., National Taiwan Univ., Taiwan, 2017.
  11. Shang-Heh Pan, ”Task-oriented Upper Limb Rehabilitation through EMG Sensing-based Intention,” Master thesis, Dept. Elect. Eng., National Taiwan Univ., Taiwan, 2017
  12. Heng-Yi Hong, ”The Circumduction Therapy on Exoskeleton Rehabilitation Robot,” Master thesis, Dept. Elect. Eng., National Taiwan Univ., Taiwan, 2016.
  13. Wei-Ming Lien, “ Developing a Novel Bilateral Arm Training on Rehabilitation Robot NTUH-II for Neurological and Orthopedic Disorders,” Master thesis, Dept. Elect. Eng., National Taiwan Univ., Taiwan, 2015.
  14. 林佳勳(Chia-Hsun Lin), “針對五十肩病患之穩定性模型及力矩增益值動態調整方法應用於NTUH-II復健機器手臂 Dynamic stiffness model and torque gain adjustment method implemented on ntuh-ii robot arm for frozen shoulder rehabilitation,” Master thesis, Dept. Elect. Eng., National Taiwan Univ., Taiwan, 2014.
  15. Guan-De Lee, “Design an assistive control system for upper limb rehabilitation robot,” Master thesis, Dept. Elect. Eng., National Taiwan Univ., Taiwan, 2012.
  16. Li-Chun Hsu, “A gravity compensation-based upper limb rehabilitation robot,” Master thesis, Dept. Elect. Eng., National Taiwan Univ., Taiwan, 2011.
  17. Bing-Chun Tsai, “An articulated rehabilitation robot for upper limb physiotherapy and training,” Master thesis, Dept. Elect. Eng., National Taiwan Univ., Taiwan, 2010.
  18. Yen-Yu Chou, “Design and application of the rehabilitation robot for upper limb physiotherapy and training,” Master thesis, Dept. Elect. Eng., National Taiwan Univ., Taiwan, 2009.

Dissertation

  1. 王威文(Wei-Wen Wang), “慣性感測系統結合外骨骼式機器人針對中風患者進行雙側上肢訓練 Bilateral arm training (BAT) using an exoskeleton robot with IMU measurement system for stroke patient,” Ph. D. Dissertation, Dept. Elect. Eng., National Taiwan Univ., Taiwan, 2014.