Advanced dementia prediction and detection via hybrid convolutional neural networks on MRI data

Alzheimer's disease (AD) is a progressive neurodegenerative disorder affecting millions worldwide, characterized by cognitive decline and memory impairment. Early detection and accurate diagnosis are critical for effective intervention and treatment planning. This research proposes a novel hybrid convolutional neural network (CNN) architecture combining CNNs with Long Short-Term Memory (LSTM) networks for enhanced dementia prediction and classification using magnetic resonance imaging (MRI) datasets. The study utilizes the comprehensive ADNI and OASIS datasets, implementing advanced preprocessing techniques including skull stripping, normalization, and tissue segmentation to extract white matter (WM) and gray matter (GM) features. The proposed hybrid model achieves superior performance with 96.2% accuracy, 95.8% precision, 95.5% recall, and 95.6% F1score in classifying five dementia stages: Cognitive Normal (CN), Significant Memory Concern (SMC), Early Mild Cognitive Impairment (EMCI), Late Mild Cognitive Impairment (LMCI), and Alzheimer's Disease (AD). Comprehensive evaluation using ROC curves, confusion matrices, and comparative analysis demonstrates the model's superiority over traditional CNN approaches and state-ofthe-art methods, establishing its potential for clinical applications in early dementia detection and staging.