Virtual Booth

This research introduces an AI-powered system that classifies dementia into four stages: non-dementia, very mild, mild, and moderate, using MRI images. Models such as VGG16, Xception, EfficientNet, and ResNet analyze hypothalamic features to indicate disease severity. A user-friendly web interface allows users to upload MRI scans and instantly receive classification results, supported by Grad-CAM heatmaps that visualize the AI’s decision process. Achieving around 90% accuracy with VGG16, this project demonstrates how deep learning can enhance medical transparency, accessibility, and early diagnosis, bridging the gap between complex AI technology and real-world clinical application.