A Novel Deep Convolutional Neural Network for Alzheimer's Disease Prediction using MRI Scans

Alzheimer's disease (AD) is a progressive neurological disorder that worsens over time and affects cognitive function and memory, making it difficult to carry out daily tasks. The buildup of amyloid plaques and tau protein tangles in the brain leads to the demise of brain cells and the disruption of neuronal connections, ultimately causing the disease. Common symptoms of AD include memory loss, confusion, and difficulty communicating. The suggested improvement entails employing a deep CNN for the automated extraction of features from MRI scans. Deep learning models have achieved astounding results in a variety of picture categorization tasks in recent years. In my research, I examine how the cutting-edge EfficientNetB0 architecture could be used to achieve high-performance picture classification. EfficientNetB0 has gained significant attention due to its efficient and effective design, making it suitable for resource-constrained environments without compromising accuracy. The proposed approach combines the power of EfficientNetB0 with global average pooling and a dense layer to predict class labels. By leveraging the pre-trained weights of EfficientNetB0, the model benefits from the learned representations of a diverse image dataset, enhancing its ability to generalize to unseen data. Fine-tuning is employed to adapt the model to the image classification task. Results showed that the suggested method worked well, with an overall accuracy of 97% across four courses. This research makes a valuable contribution to the domain of image classification by showcasing the capabilities of the EfficientNetB0 architecture and its potential to achieve high-performance results.