Abstract: The present disclosure describes a system, method and software applications for predicting the KAM of a subject for a plurality of gait cycles of a subject. At least one sensor attachable adjacent to the ankle of at least one leg of the subject provides accelerometer and gyroscopic data for the plurality of gait cycles. A processing means receives subject parameters and accelerometer and gyroscopic data from the at least one sensor for evaluation by a neural network of predicted KAM values for the plurality of gait cycles of the subject. The neural network is configured via determining KAM values for each of the plurality of subjects from measurements during a plurality of gait cycles and other parameters.

Abstract: The present disclosure describes a system, method and software applications for predicting the KAM of a subject for a plurality of gait cycles of a subject. At least one sensor attachable adjacent to the ankle of at least one leg of the subject provides accelerometer and gyroscopic data for the plurality of gait cycles. A processing means receives subject parameters and accelerometer and gyroscopic data from the at least one sensor for evaluation by a neural network of predicted KAM values for the plurality of gait cycles of the subject. The neural network is configured via determining KAM values for each of the plurality of subjects from measurements during a plurality of gait cycles and other parameters.