Alzheimer’s Disease Image Diagnosis by Active Contour Segmentation &Bootstrap Bagging Learning Model

Abstract

AD is a neurodegenerative disease in which pathological changes occur decades before disease manifestation. The disease is characterized by the formation of senile plaques, NFTs and subsequent synaptic loss and neuro degeneration. Although AD affects a major part of the population worldwide, to date, there is no therapy to cure AD. Since disease-modifying therapies may be the most beneficial in early stages of the disease, it is important to diagnose AD as early as possible. This paper proposed a model of disease diagnosis and identify the class of input image into healthy, mild, severe. Input image was preprocessed in order to extract brain region by use of active contour method. Further image was break down into blocks and estimate the histogram features. Extracted features were used for the training of bootstrap bagging learning model. Experiment was done on real image dataset of Alzheimer’s disease of all set of classes. Result shows that proposed Alzheimer’s disease Class Prediction by BioGeographic Optimization & Bootstrap Model (ADCPBOBM) model has improved the work detection accuracy of correct.