Early Prophase of Alzheimer Disease Using Deep Learning

Abstract

Alzheimer's disease is a neurological disorder that mostly affects people around age 60 and above. Some cases of AD are found in people around age 40 to 60. This case is called the early cause of Alzheimer's disease. AD is the main reason for dementia in elderly people which causes memory loss. Loss of synapsis in the brain is the cause of AD, synapsis is the gap or separation between neurons, a tissue in the brain which takes care of all the functions that the brain does. Synapsis is responsible for making the connection between neurons which stores memory that we see, do or hear daily. When this synapsis is lost causes memory loss dementia which in turns severe are called Alzheimer's disease. AD medical disorder is due to various reasons but the most common reason is the accumulation of amyloid-β in the separation of neurons which causes the loss of synapsis. The more clumps amyloid-β formed in the separation the more the severity of case and the loss of synapsis. In early stages there are very few losses of synapsis which cannot be traced by a doctor, if AD is identified at its early stage its progress can be reduced by reducing the amount of amyloid-β accumulating the brain thus reducing the effort of Alzheimer's disease. Modern scan technologies have proven accuracy in brain imaging which can be used for prediction of AD. Computer vision and deep learning techniques can be used to predict AD from these MRI scan images. CNN deep learning model which is specifically used for classifying images multiple combinations can be used with CNN to fit it to the use case. We picked CNN and obtained the scan images dataset. The model is trained using the dataset. The implementation is done using python on google code laboratory.