Motor disabled patients usually use their residual motor functions to control welfare apparatus. It is desired to develop the control command input system which can be applied to the patients with various paralysis condition. In this paper, the command input system which consists of a 3 D position measuring device and an artificial neural network (ANN) for detecting control commands from residual motor functions of the patients was studied. The ANN detects a specific movement by the patients to decide a control command. The control command is selected by the change of 3 D position which is caused by the same part movement as the specific movement. Considering the clinical application of the command input system with C 4 quadriplegic patients, recognition of the nodding by the ANN was studied at first. Then, the command input system which includes the nodding detection and command selection by head movements was examined in character input with four normal subjects and a C 5-6 quadriplegic patient. In application of the command input system to a C 5-6 quadriplegic patient who has parts of residual motor functions of upper extremities, change of hand position on a desk by rapid elbow flexion were recognized by the ANN to decide a control command. The performance of the command input system which uses the rapid elbow flexion for the decision and the change of hand position on the desk for the command selection was examined through character input experiments. From those examinations, the command input system was found to be applicable to different parts of residual motor functions providing good performance of system operation. The experimental results also suggested that the control command source have to be selected appropriately on a patient. Therefore, this system can be useful for motor disabled patients with different paralysis condition.
|Number of pages||9|
|Journal||japanese journal of medical electronics and biological engineering|
|Publication status||Published - 1999 Jul 17|
ASJC Scopus subject areas
- Biomedical Engineering