DeepCurvMRI: Deep Convolutional Curvelet TransformBased MRI Approach for Early Detection of Alzheimer’s Disease

Abstract

Alzheimer’s disease (AD) is one of the most prevalent neurodegenerative disorders worldwide, affecting millions of individuals and imposing a significant burden on healthcare systems. Early and accurate diagnosis remains a critical challenge, as clinical assessment alone is often insufficient to detect subtle neurological changes in the initial stages. This paper presents NeuroScan AI — an intelligent, web-based MRI classification system that leverages ensemble deep learning to automate the diagnosis of Alzheimer’s disease from brain MRI scans. The system integrates two powerful Convolutional Neural Network architectures — MobileNetV2 and EfficientNetB0 — and combines predictions through four ensemble strategies: Average Probability, Majority Vote, Weighted Average, and Max Probability. Trained and evaluated on 17,287 MRI images spanning four clinically defined dementia stages, the system achieves up to 99.2% classification accuracy. The platform is deployed as a secure Flask web application with role-based access control, administrator management panel, and a real-time prediction dashboard for medical professionals.