A Rehabilitation Guidance System using Multi-Layer Feed-Forward Neural Networks

K. Keeni, W. Yan, N. Kitazima, and T. Takeda (Japan)


Rehabilitation, Time-series prediction, Neural networks,Back propagation


This study focuses on the subject of guiding a subject tak ing rehabilitation exercise. A simple scenario where a sub ject rides on a bicycle for performing rehabilitation exercise is considered in this study. The system guides the subject in three ways. Firstly, it reschedules the Heart Rate (H/R) chart from the H/R chart prescribed by a physician, sec ondly it guides the subject about the ideal paddling speed, and finally it instructs the subject about when to accelerate, decelerate or hold the present paddling speed. Two feed forward neural networks are employed to deal with these problems. The result achieved from the preliminary part of this study is promising.

Important Links:

Go Back