An application of Neural Network for paraplegic quadriceps muscle model

This paper presents the development of quadriceps muscle model based on Functional Electrical Stimulation (FES). Artificial Neural Network (ANN) were used to study the impact of stimulation frequency, pulse width and pulse duration towards the output torque produce by paraplegic.722 data are used by randomly divide 70% for training, 15% for validation and another 15% is for testing process. Two types of training approaches which is Levenberg Marquardt Back propagation (LM) and Resilient Back propagation (RP) are used in developing of quadriceps muscle model. The model developed are validate with the clinical data to see the accuracy of the torque output predicted with the identified parameter. From the study, LM is found to be the most accurate with accuracy up to 99.98% The identified parameter used from model developed in this study will be used to control various strategies on the Functional Electrical Stimulation (FES) system.