In our research, we applied novel Graph Signal Processing techniques and machine learning to EEG data for Alzheimer Disease (AD detection. Evaluating several models on the OpenNeuro dataset ds004504, we achieved up to a 85% accuracy for binary classification of AD. Our methods presents a framework for further research, potentially extending to other neurodegenerative diseases and different data types, like neuroimaging in various disorders. For more information and code, please the GitHub page here.
