This project applied graph neural networks (GNNs) to analyze MRI brain images for Alzheimer's disease classification. Faced with limited original data, I enhanced the dataset using methods tailored for graph data and adapted techniques from computer vision. These efforts resulted in a classifier accuracy improvement of over 10%.