Neural Network with Improved Gradient Amplitude (NN-IGA) for Detecting and Diagnosing the Cerebral Palsy (CP)

Abstract

Cerebral palsy (CP) is set of disorders that can affect the human ability to progress and preserve challenge and posture. This is basic disability that can occur in childhood. Based on the ability of the brain this can damage the eye of the human. Children between ages 3-11 years can effect with this disorder. This can capture by observing the movement of the eyes of the children. The effected children will have the mild motor-impairment and also we have analyzed the performance of CP children periodically. To decrease the risk of this disease, there is a need of developing the automated learning algorithm to detect the cerebral palsy (CP) in kids very effectively. In this paper, Neural Network with Improved Gradient Amplitude (NN-IGA) is developed to overcome the various issues in detecting the abnormalities in CP for kids and the performance is improved with more accuracy. To get the more accuracy the canny edge detection is used to overcome the issues in finding the edge detection. The comparative results shown among random forest (RF), neural network with canny and NN-IGA. The proposed system NN-IGA shows the more accuracy when compare with existing approaches.