Dynamic Simulation for Ergometry Cycling in Rehabilitation

Every year, thousands of people suffer from spinal cord injuries due to accidents or falls. the effects can be very upsetting, resulting in paralysis or loss of sensation. this sudden loss of mobility and senses causes significant stress on patients both physically and psychologically, such as demineralisation of bones, frequent skin problems pressure sores, increased incidence of urinary tract infection, impaired lymphatic and vascular capacities. apart from the broken nervous connection, the joints are still physically connected. this has prompted engineers and physicians to come up with idea of artificially stimulating the muscles of paraplegics to produce certain desired leg movements, a technique commonly known as functional electrical stimulation (fes). the most common method of fes is to apply electrodes to the skin over the muscle and control the stimulation by the variation of potential and frequency. the first priority for rehabilitation engineering is to enhance the paraplegic's health, which can be achieved by applying fes on the large muscles of their legs to exercise them. this type of exercise should be both enjoyable and convenient. one of the most promising exercises is cycling. these days fes has been used widely for cycle ergometry to reduce spasticity, reverse muscle atrophy, as well as increase cardio vascular fitness, muscle bulk, blood flow and bone density in the legs, and to reduce the chance of developing pressure sores and cardiovascular disorders such as orthopaedic intolerance. cycling by means of fes is an attractive training method for spinal cord injured (sci) subjects. fes-cycling performance is influenced by a number of parameters like seating position, physiological parameters, conditions of surface stimulation, and pedaling rate. cycling using fes is also a very promising approach because it activates most of the muscle groups in the lower limbs under safety conditions. for this reason several methods of stimulation have been used according to the activated muscle groups, allowing for different control strategies. to obtain the desired leg motion, the stimulation is embedded into a control system, where the stimulator is controlled by a microcomputer. measurements could then be made to give feedback to the control system so that the performance could be further improved. this paper presents the development of a strategy for cyclical motion control by controlling the torques of the lower limb joints. a humanoid model is developed in this study and used as a test-bed to develop a suitable strategy for motion control of leg joints to result in bipedal motion in the lower extremity, and to identify torque profiles in each joint in different bipedal sitting positions.