Alzheimer Disease Detection using Convolutional Neural Network and Particle Swarm Optimization

Alzheimer's disease (AD) is a major global health concern, particularly with aging populations. Early detection and diagnosis play a critical role in managing the progression of the disease. Advances in neuroimaging have enabled the use of 3D brain images, which offer valuable insights into disease biomarkers. This paper proposes a specialized Con-volutional Neural Network (CNN) models for interpreting 3D brain MRI scans in the context of Alzheimer's disease detection. By training distinct models for the axial, coronal, and sagittal orientations, we aimed to capture the unique anatomical features and variations inherent to each plane, which enhanced feature extraction and classification accuracy. Our approach contrasts with traditional unified models by focusing on optimizing the individual contributions of each orientation. Furthermore, we employed a Particle Swarm Optimization (PSO) algorithm to assign coefficients to each model, enabling a refined integration of their outputs. This approach facilitated a notable improvement in overall performance, reflecting the advantages of combining specialized CNN architectures with optimization techniques for medical image analysis.