A Robust Deep Neural Network Framework for the Detection of Dementia

The basic goal of this Deep Learning (DL) framework is to identify early pathological signals that people could miss in order to improve physicians' effectiveness in disease identification. The most common degenerative condition that causes dementia, is the Alzheimer's disease (AD), it offers serious health hazards to middle aged and older people. Different people can experience dementia in different ways and it will create individual signs dependent on the fragment of their brain which is affected. However, it is more essential to treat this disease because, the person with this disease may suffer to communicate with others, to eat and chew their own food etc. Several state-of-art methods are utilized for the detection of dementia earlier. In this study, the detection of dementia is done by DL based Convolutional Neural Network (CNN). The Grey Level Co-occurrence Matrix (GLCM) is recycled for the feature extraction process. The Spider Monkey Optimization (SMO) is used in this study for feature selection process. Subsequently, CNN is a deep learning (DL) network architecture that it will learn directly from the data. This will be utilized for identifying objects and the category of the image pattern. Moreover, this can also be used for signaling the data, time-series and categorizing signal. Therefore, the investigational outcome reveals that the future algorithm has the ability to identify and treat dementia. The effectiveness of the proposed work can be proved by comparing the CNN with other existing models. The comparison outcome gives the proposed system has better accuracy than other existing methods.