A study on a new control command input device for functional electrical stimulation system - 2nd report: improvement of control command input method using neural network

摘要

In order to detect the head movements as a control command input device for the functional electrical stimulation (FES) system, a gyroscope was used. The angle of the head movements to be detected by the gyroscope, however, had some errors because of its frequency characteristic. This paper described that an artificial neural network was applied to the control command input device for FES system. The neural network was learned by teacher signals which were obtained by the image analysis. To evaluate the input device using the neural network, target tracking experiments were done. From the results, it was cleared its operationality was improved.