Intelligent Haptic Robotic Device for Upper-Limb Stroke Patients
Keywords: Haptic, stroke, automated, rehabilitation.
In collaboration with: Quanser and Toronto Rehabilitation Institute.
Overview of Research
Hemiparetic stroke survivors with an affected upper limb (i.e. arm) usually have great difficulties performing many activities of daily living (ADL), such as pushing on a chair to stand up, getting into bed, or dressing themselves due to a lack of coordination and weakness in muscle control. Therapists work with stroke survivors to improve coordination and enable more muscle control in the arm in order to decrease impairment and increase functionality.
In the early stages of therapy, stroke survivors often have very little control of their limb and require extensive practice and assistance to regain motor skills. Consequently, therapists must spend significant amounts of time guiding stroke survivors through exercises and functional activities, which can be difficult for both. The goal of this research is to develop an intelligent haptic robotic rehabilitation system to augment treatment of the upper limb post stroke. The block diagram in Figure 1 depicts the overall intelligent haptic robotic system, representing three research and development focus areas, consisting of a hardware device, an artificial intelligence controller, and a virtual reality user interface.
Figure 1: Block diagram of AUSR system - red dashed box denotes
hardware/software boundary (click on figure to enlarge)
In the first project, discussions with experienced stroke therapists identified early stage exercises for upper limb stroke survivors as an area of rehabilitation that is in need of supplementary treatment tools. An upper limb stroke rehabilitation prototype was designed to assist therapy in the early stages of stroke recovery. The goal was to simulate targeted linear-reaching task therapy, using elbow position monitoring and assistance to rehabilitate reaching motor skills. The device was intended to autonomously adjust to meet the specific abilities of different patients, as well as increase the difficulty of the exercise as patient performance improves. Feedback was to be provided to both the patient (in the form of graphic motivation and games) and therapist (in the form of patient performance statistics).
The first robotic prototype was completed and is shown in Figure 2. The non-restraining, haptic platform provided resistance and directional guidance for the user during the exercise. It had two active and two passive degrees of freedom, and allowed the reaching exercise of either upper-limb to be performed in three-dimensional space. Built-in encoders and unobtrusive sensors provided data to indicate hand position and abnormal upper-limb posture during the rehabilitation exercise, respectively. Other data, such as velocity and force, could be calculated to monitor proper technique and posture. This array of information formed the input parameters for the controller.
Figure 2: Upper-limb stroke rehabilitation prototype
The control system was based on a partially observable Markov decision process (POMDP), a versatile decision-theoretic modelling technique, and was responsible for guiding the patient through the reaching exercise. It could autonomously adjust the exercise parameters (e.g. resistance, target distance) to the abilities of each patient and increase the difficulty of the exercise as the patient improves. The goal of the controller was to help the patient achieve their maximum reaching distance at the greatest level of resistance, while maintaining control. Figure 3 depicts the current POMDP controller, modelled as a dynamic Bayesian network (DBN).
Figure 3: Diagram of POMDP controller (click on figure to enlarge)
Videos
The following video demonstrates the task the AUSR system is being designed to replicate (right-click and "Save as..."). If you would like a copy of the video in another format than .wmv, please e-mail Jennifer Boger.
Current Research
Robotic System
The redesigned robotic device has been developed in collaboration with Quanser Inc. It has been designed using our experiences with a previous prototype and an international survey of therapists.
Device Description
The upper limb stroke rehabilitation robot prototype is a two degrees of freedom (2DOF) impedance controlled planar haptic robotic device (see Figure 4a). The robot’s casing dimensions are approximately 32x14x39cm and its weight is about 17.3kg.
Figure 4: (a) Haptic robotic device and (b) range of motion.
Two DC motors attached with optical encoders drive the robotic arm. The robotic arm is connected to a plastic end-effector on a caster wheel which stabilizes the end effector on a flat surface. The maximum force that can be applied on the end-effector is 52.8N per motor. The optical resolution at the end effector is 0.013 mm/count at home position. This end effector is customizable for different patient requirements. Figure 4b shows the robot workspace created with its full range motion. The workspace is an irregular shape having an approximate dimension of 93 cm x 35 cm.
The robot is equipped with an external emergency stop to shut off power for safety purposes. It has a USB interface with a computer which makes it possible to operate the robot with a laptop or tablet device.
The robot is operated by QuaRC (Quanser’s Rapid Control Prototyping) software and it can be used with other programming languages using standard communication protocols, e.g., TCP/IP and shared memory.
Graphical user interface (GUI)
Figure 5 shows the GUI developed for the robotic system.
Figure 5: Haptic robot's (a) control interface and (b) visual feedback.
Figure 5a depicts the control interface of the robot and Figure 5b shows a 3D viewer for real-time visual feedback of an exercise, where the robot end-effector is shown using a red ball and the workspace is shown as a table top. The GUI includes therapist interfaces for both manual and automatic haptic exercises. For each exercise it shows exercise information for performance evaluation. It also provides games that employ VR for encouraging the patients to be engaged in the repetitive motion exercises.
Video games
The GUI also includes video games to encourage the patient to perform repetitive motion exercises. Figure 6 shows three video games, namely, Bouncing ball, Object rearranging, and Car driving games.
Figure 6: Various games that can be played with the robot. (a) Bouncing ball game, (b) object rearranging game, and (c) car driving game.
In the bouncing ball game (Figure 6a), the patient controls a bat position with the robot end-effector to prevent a bouncing ball from hitting the ground.
The object rearranging game (Figure 6b) includes some objects common to a living room. The objects are initially placed randomly in the environment. The patient can control a cursor with the end-effector to select and reposition an object in the environment. The overall goal of the game is to rearrange all the objects so that the environment looks like a familiar living room environment.
In the car driving game (Figure 6c)the patient controls the speed and position of the car to avoid collisions with the obstacles placed on the road. The video games provide several options to change the difficulty levels to match individual performance. Different haptic effects are applied to improve muscle activities. The game GUI also provides scores so that the patient is encouraged to play the game for longer duration.
Publications
- Huq, R., Wang , R., Lu, E., Lacheray, H., and Mihailidis, A. (in review, 2012) Development of a Fuzzy Logic Based Intelligent System for Autonomous Guidance of Poststroke Rehabilitation Exercise. IEEE/ASME Transactions on Mechatronics.
- Huq, R., Lu, E., Wang, R., and Mihailidis, A. (in review, 2012). Development of a Portable Robot and Graphical User Interface for Haptic Rehabilitation Exercise. In the IEEE International Conference on Biomedical Robotics and Biomechatronics, June 24-28, Rome, Italy.
- Lu, E., Wang, R., Huq, R., Gardner, D., Karam, P., Zabjek, K., Hébert, D., Boger, J., and Mihailidis, A. (in press, 2012). Development of a robotic device for upper limb stroke rehabilitation: A user-centered design approach. Journal of Behavioral Robotics.
- Huq, R., Lu, E., Wang, R., and Mihailidis, A. (in review, 2011) Development of a Portable Robot and Graphical User Interface for Haptic Rehabilitation Exercise. In the 2012 Haptics Symposium, March 4-7, Vancouver, Canada.
- Kan, P., Huq, R., Hoey, J., Goestschalckx, R., Mihailidis, A. (2011) The development of an adaptive upper-limb stroke rehabilitation robotic system. Journal of Neuroengineering and Rehabilitation, 8(33). doi:10.1186/1743-0003-8-33.
- Huq, R., Kan, P., Goetschalckx, R., Hébert, D., Hoey, J., and Mihailidis, A. (2011). A Decision-Theoretic Approach in the Design of an Adaptive Upper-Limb Stroke Rehabilitation Robot. In the International Conference of Rehabilitation Robotics (ICORR), June 29 - July 1, Zurich, Switzerland, pp. 589-596.
- Lu, E., Wang, R., Hebert, D., Boger, J., Galea, M., and Mihailidis, A. (2011). The development of an upper limb stroke rehabilitation robot: Identification of clinical practices and design requirements through a survey of therapists. Disability and Rehabilitation: Assistive Technology, 6(5), 420-431.
- Kan, P., Hoey, J. and Mihailidis, A. (2008) Automated upper extremity rehabilitation for stroke patients using a partially observable Markov decision process. In AAAI 2008 Fall Symposium on AI in Eldercare: New Solutions to Old Problems, Arlington, VA.
- Kan, P., Boutilier, C., Hebert, D., Hoey, J., Boger, J. and Mihailidis, A. (2007). The preliminary development of a POMDP controller for upper-limb stroke rehabilitation. Festival of Intl Conf on Caregiving, Disability, Aging and Technology (FICCDAT): The 2nd Intl Conf on Technology and Aging (ICTA), Toronto, Canada.
- Lam, P., Hebert, D., Boger, J., Lacheray, H., Gardner, D., Apkarian, J. and Mihailidis, A. (2008). A haptic-robotic platform for upper-limb reaching stroke therapy: Preliminary design and evaluation results. Journal of NeuroEngineering and Rehabilitation, 5(15). doi: 10.1186/1743-0003-5-15.
Funding Sources
- Communications and Information Technology Ontario (CITO)
- Precarn Incorporated
- NSERC-CIHR Collaborative Health Research Program grant
Industrial Partner
Research Team
Alex Mihailidis, Ph.D., P.Eng. (University of Toronto)
Elaine Lu, B.Eng, M.A. (University of Toronto)
Rajibul Huq, Ph.D., P.Eng. (University of Toronto) Rosalie Wang, B.Sc.(OT), PhD (Toronto Rehabilitation Institute)
Jacob Apkarian, Ph.D. (Director R&D, Quanser)
Debbie Hébert, M.Sc. (Occupational Science and Occupational Therapy, University of Toronto)
Craig Boutilier, Ph. D. (Chair, Computer Science, University of Toronto)
Jesse Hoey, Ph. D. (Computer Science, University of Toronto)
Geoff Fernie, Ph.D. P.Eng. (Toronto Rehabilitation Institute)
