The main aim of this project is to develop and validate stimulation patterns for walking of hemiplegic patients assisted by Functional Electrical Stimulation (FES). These patterns are the prerequisite for the rule-based control of FES assisted walking. The design of rules for control comprises the following two steps: 1) determination of muscle activation patterns by using a customized spatial (3D) model of a hemiplegic person walking, and 2) acquiring the rules, that is, determining the correlations between the muscle activation patterns and kinematics of walking by means of artificial neural networks. The adopted FES system is planned to include up to four channels of stimulation and apply surface electrodes. The use of multipolar electrodes is anticipated, allowing selective activation and, even more important, control of the stiffness of the joints in question. The joints allowing movement in the coronal plane are the hips, and joints in the sagittal plane are ankles, knees and hips. The simulation profiles will be determined by minimizing the tracking error of the joint angles and optimizing the joint stiffness that leads to minimization of risk of falling. A radial-basis function artificial neural network will be applied for acquiring the rules that will be used for real-time control. The major effort at the beginning of the project will be to refine the existing dynamic models used for walking and develop a computer interface that allows easy and effective testing of various strategies. A new simulation study involves development of a new 3D walking model which consists of the following actions: 1) assuming a suitable multi segment model with appropriate number of degrees of freedom, 2) developing equations for 3D dynamics of walking, 3) determining suitable set of input parameters which are user specific and easily measurable, 4) assuming suitable muscle models for measuring muscle torque, and 5) assessing parameters of the model. The results of the simulation studies will be validated by comparing the muscle activation patterns obtained as outputs of simulation studies with the experimental EMG recordings and literature data. The work on the simulation will be in parallel with the experiments examining the kinetics and EMG in able-bodied subjects during the following tasks: walking, gait initiation and gait termination. Able-bodied subjects will be instrumented with surface EMG electrodes at ankle flexors and extensors, knee flexors and extensors, hip flexors and extensors and hip abductors and adductors to record muscle activation patterns during the above tasks. Subjects will also be instrumented with Penny and Giles goniometers to measure ankle, knee and hip joint angles, sensors to measure hip acceleration, and pressure sensitive sensors in the shoes to estimate ground reactions.
|Effective start/end date||31/12/2004 → 31/12/2004|