B Santhosh Kumar, P. Penchala Prasad, M. Raghavendra Reddy
Abstract Alzheimer’s disease is a neurodegenerative disorder that affects millions of individuals worldwide, making early diagnosis through Magnetic Resonance Imaging a significant clinical necessity. Existing medical image analysis techniques often suffer from limitations associated with inadequate preprocessing, reduced sensitivity to subtle abnormalities in the hippocampus and cortex, poor generalization across heterogeneous MRI acquisition systems, and insufficient mechanisms for secure medical data management. To address these challenges, this research proposes an integrated framework combining the Internet of Medical Things (IoMT), Artificial Intelligence, and blockchain technology for secure and efficient Alzheimer’s disease monitoring. The proposed framework employs Feature Pooling VGG16 (FPVGG16) for discriminative feature extraction, while feature selection is optimized using the Wave Search Binary Waterwheel Plant Optimization algorithm. Subsequently, a feature-selective Coordinated Xception-based Convolutional Spatial Network (CXCSN) is utilized for accurate disease classification. Blockchain technology is incorporated to provide secure, tamper-resistant, and privacy-preserving management of patient information and MRI records. Experimental evaluations conducted on the Alzheimer’s Disease Neuroimaging Initiative (ADNI) and Open Access Series of Imaging Studies (OASIS) datasets validate the effectiveness of the proposed framework, achieving accuracies of 99.31% and 99.21%, precisions of 99.28% and 99.25%, and recalls of 99.18% and 99.14 %, respectively. The results indicate that the proposed framework provides an effective solution for secure, reliable, and highly accurate Alzheimer’s disease diagnosis and monitoring.