Prediction of sagittal lower limb joints moments under dynamic condition: feasibility of using EMG and ARMA model identification techniques

A novel alternative method reducing the need for direct inverse dynamics to solve the muscle redundancy problem at human lower limbs is proposed. It aims at computing lower limb joints moments under dynamic conditions using only electromyographic (EMG) signals in combination with an auto regressive moving average (ARMA) model. The experimental protocol is conducted by practicing full gait cycle trials in an effort to calculate joint moments. Quantitative comparisons with the output of a biological-based model showed that the proposed method is able to: 1) produce accurate estimates of the resultant moment; 2) maintain the obtained accuracy regardless of the information about status of the angle or its derivatives. The joint moment prediction by the ARMA model attained an average of R2 = 1.73. The model is characterised by stability, accuracy and minimum number of input variables. These characteristics represent an added value to be utilised in lower limbs rehabilitation.