Unveiling Diagnostic Potential: EEG Microstate Representation Model for Alzheimer's Disease and Frontotemporal Dementia

Alzheimer's disease (AD) and frontotemporal dementia (FTD) represent distinct neurodegenerative disorders affecting the brain. AD, characterized by beta-amyloid plaques and tau protein tangles, primarily impacts memory-related brain regions, leading to progressive cognitive decline and daily task impairment. On the other hand, FTD involves abnormalities in Tau or TDP-43 proteins, affecting personality, social behavior, language skills, and executive functions. Hence, early diagnosis of AD or FTD is crucial for effective management, which encompasses appropriate medical treatments, therapy and care services, social support, as well as environmental adjustments. Resting-state EEG microstates especially reflect transient brain networks, providing insights into spontaneous consciousness and aiding in diagnosing neurological disorders. We propose a novel EEG microstates-based representation model to validate it as a potential diagnostic biomarker for AD/FTD. By learning representations from EEG microstate sequences, we lay the groundwork for future effective deep-learning methods by leveraging this information.