Classifying the stages of Alzheimer's disease by using multi layer feed forward neural network

Alzheimer's disease is a type of physical brain disorder. Once the disease has begun, it is unrealistic to stop the progression. However, advanced prediction can help to diminish the functions of proteins that influence the disease. Many authors used machine learning techniques to diagnose the disease at its early stage but failed to achieve accurate classification accuracy. So in this work, we attempted to build a model that is a combination of a deep neural network and a multi-layer feed-forward neural network that could accurately extract features from input and achieve the best result. The proposed hybrid model is implemented with a specific configuration for feature extraction and classification, and this model experiments on the MRI dataset collected from ADNI and Kaggle datasets. Experimental results show that an average proposed approach achieves 93.38% of accuracy, 98.5% of AUC 90.05% sensitivity, and 95.04% of specificity for multi-classification between AD, MCI, and CN.