Gurur Gamgam has successfully defended his PhD thesis entitled "Geometric Deep Learning of Brain Connectomes Over the Spectrum of Dementia" on January 13, 2025. Dr. Gamgam studied the application of Graph Neural Networks to brain connectome analysis. He proposed the novel Structure-Function Discrepancy score extracted from a subjects structural and functional connectomes simultaneously using a Siamese GNN architecture. Experimental studies showed that the Structure-Function Discrepancy increases with progressing cognitive decline from Subjective Cognitive Impairment to ADD. He also showed that multi-head GNNs with a targeted disentanglement give promising results in locating the subnetworks that undergo changes. Congratulations Dr. Gamgam.