Rehabilitation Approach Based on Parallel Neural Networks
10.2185/jjrm.56.34
- Author:
Akifumi KOUMURA
;
Shinichi YOSHIDA
- Publication Type:Journal Article
- MeSH:
Fingers, unit of measurement;
Neural Network Simulation;
Rehabilitation aspects;
Arm;
Approaches
- From:Journal of the Japanese Association of Rural Medicine
2007;56(1):34-38
- CountryJapan
- Language:Japanese
-
Abstract:
Basal ganglia are an area where cerebral haemorrhage occures frequenthy. Recently our knowledge of the brain and its function increases rapidly, but this knowledge is not much used in rehabilitation yet. Parallel neural networks (Hikosaka, 1999) are a model of sequential procedures learning, and basal ganglia perform an important function in it. Using this model, we designed a rehabilitation approach and applied it to a patient with left putamen hemorrhage. In the early stage, we promoted passive movement of her paralysed arm and fingers in the spatial coordinates with her eyes opened. Afterwords, we guided her to motor leaning in the motor coordinates with somatic sensation with her eyes closed. We intended to integrate the spatial coordinates and motor coordinates, by which we encouraged her to move her arm and fingers with her eyes opened after her recognition of motion perception improved. Her motor paralysis of fingers in particular improved by rehabilitation, which lasted about 1.5 month. We recognized some improvement in her arm, but it was slightly insufficient than her fingers. We considered that a difference in the degree of improvement occurred because she could not confirm enough on her shoulder in the early stage.
