A Deep Learning Based Classification of Alzheimer's Disease Using 3D CNN

Alzheimer's disease(AD) is a critical neurological disorder frequently characterized by cognitive decline and physical deterioration. Currently, a combination of clinical evaluations and psychological testing is used to diagnose AD. Though these diagnoses aren't flawless, more sophisticated diagnostic equipment can help us learn more about AD and increase our capacity to identify the condition. Early and accurate diagnosis of AD is crucial for timely intervention and personalized treatment strategies. Functional Magnetic Resonance Imaging (fMRI) offers a window into the dynamic neural activity alterations associated with AD, presenting an opportunity for computational approaches to aid in disease classification. This study provides a deep learning-based approach that uses fMRI data to classify Alzheimer's disease cases. The proposed methodology leverages the power of deep neural networks to automatically extract intricate patterns and features from fMRI scans, capturing subtle functional alterations indicative of AD. This work employs a multi-layered 3-dimensional convolutional neural network (CNN) architecture to efficiently learn hierarchical representations from spatial and temporal characteristics of fMRI data. The model is trained on the Alzheimer's Disease Neuroimaging Initiative(ADNI) dataset, incorporating both healthy control subjects and individuals diagnosed with Alzheimer's disease. The performance of the suggested deep learning-based method is thoroughly evaluated using a comprehensive set of metrics such as sensitivity, specificity, and accuracy. Comparative analyses are conducted against traditional machine learning approaches and state-of-the-art methods to demonstrate the superior classification performance and efficiency of the proposed model.