A New Concept in the Evaluation of Cybernetic Actuators Control Using Virtual Reality

In order to control the movement of a cybernetic actuator the EMG signal is generally used as a source of command This signal has to be processed in order to extract relevant features, which are then classified. Many schemes exist today in both feature extr action and classification, each one claiming to reduce the error rate and there has been some approaches in order to assess the input-output characteristics of prostheses. This paper vvill introduce a new concept in developing a unique platform using virtual reality (VR) tools for evaluating both the different schemes of EMG signal processing and cybernetic control. The design follows a modular approach allowing for the change of each module (analog signal conditiorming, data acquisition, signal processing, actuator control, VR aspects) accordingly to the specific needs of an application. The foreseen applications of this work are performance evaluation of EMG signal processing algorithms for prosthesis control in real conditions, performance evaluation of the motor control schemes by executing real tasks, selection of the optimum scheme for a particular application (spatial, medical surgery, underwater, etc...) and training of amputees or future user s of the system in real conditions where "real conditions" means the VR simulated environment. INTRODUCTION With the price reduction of new technologies, it is now possible to apply these technologies to rehabilitation engineering. This field has unfortunately not benefited from the results of advances in such growing fields as VLSI electronics compared to other ones although simple robotic devices have been used for physical therapy and rehabilitation [1] One of these new technologies is virtual reality (VR): two years ago, only an organisation could afford such a system (with prices from around $100 000 US for an acceptable system), but it is now possible even for the individual to purchase his own PC based VR sytem for less than a thousand $ US. On the other hand, with the tremendous advances in microelectronics, it is also possible to design smart myo-electric prostheses,. This new generation of artificial devices will be fitted per sonally to each patient, taldng into account both mechanical (e. g.. socket fitting to the stump) and electrical aspects. As the electromyogram (EMG) signal is generally the best candidate and used by many researchers [2]-[4], the latter aspects consist essentially of EMG signal processing in order to determine which feature to extract keeping in mind that the requested characteristics for these 49 From "MEC 95," Proceedings of the 1995 MyoElectric Controls/Powered Prosthetics Symposium Fredericton, New Brunswick, Canada: August, 1995. Copyright University of New Brunswick. Distributed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License by UNB and the Institute of Biomedical Engineering, through a partnership with Duke University and the Open Prosthetics Project. University of New Brunswick 1_ features are good controlabilty by the subject., ease of training and real-time implementation, The requested characteristics for the cybernetic actuator itself are: anthropomorphism, low-weight, low-cost. There is obviously a need to evaluate the effectiveness of EMG signalprocessing and cybernetic control schemes in order to choose the most appropriate one for a particular case of rehabilitation. Although there has been some research in this field [5], unfortunately there has been no harmonizing work in order to compare in the same real conditions the performance of these algorithms when they are used in their respective environment. In the following sections, the design concept of a rehabilitation device using VR tools will be described. The basic idea is to prototype a modular system leading to a single platform allowing to implement and test different schemes of EMG signal processing and cybernetic control. Each module may consist of hardware and/or software, but all parameters should be software programmable. The modules are: analog signal conditionning, data acquisition, EMG signal processing, actuator control, VR environment generator and display,. These modules are selected accordingly to the specific needs of an application-patient combination. Both common daily and professional activities are considered. This idea is in fact the extension ofpreviously observations of the effectiveness of using computer graphics in order to automate the learning of an artificial arm [6147]. It is thought that this new approach can offer the following advantages over the classical one: selection of the best scheme (signal processing and actuator control) for a specific user profile by the defmition of an objective performance index, easy and attractive training of amputees by allowing for self-paced, self-selected environment (especially for children) and obje.ctive follow-up of the amputee Fig. 1: The set-up., METHODS The set-up is schematized in figure 1. EMG signals are picked-up by surface electrodes. The number and position of these electrodes are determined by the type of EMG signal processing. In this specific example they are to be positionned at the bulk of the biceps-brachii and triceps-brachii muscles