Introduction: Electroencephalography (EEG)-based classification of Alzheimer's disease (AD) and frontotemporal dementia (FTD) relative to cognitively normal (CN) controls is commonly interpreted through stable disease-related patterns. However, classification-relevant EEG responses in computational models may also appear fragmented or weakly preserved, making subject-level reliability difficult to assess. This study investigated whether recurrent disorder-like EEG patterns can provide exploratory evidence of subject-wise discriminative organization in dementia classification. Methods: We analyzed a publicly available resting-state EEG dataset including AD, FTD, and CN subjects. Clustered Pattern Projection (CPP) was applied to Dynamic Mode Decomposition (DMD)-based epoch descriptors to construct prototype-based EEG representations. Classification was performed using a linear support vector machine under a nested leave-one-subject-out cross-validation (LOSO-CV) framework. Subject-level reliability was further examined using margin-based analysis. Results: CPP showed competitive subject-level performance, particularly in the FTD vs. CN classification task, a setting in which resting-state discriminative patterns are often less consistently preserved than in AD. In addition, task-dependent subject-level margin patterns were observed under strict subject-wise validation, suggesting that margin analysis may provide a useful exploratory tool for evaluating the reliability of learned EEG representations. Discussion: These findings suggest that EEG generalization in dementia classification should not be interpreted only through preserved canonical biomarkers. Instead, recurrent disorder-like patterns may contribute to computationally detectable decision structure, and CPP provides a framework for examining such patterns under strict subject-wise validation.

Background/Objectives: Detecting Alzheimer’s disease (AD) from normal aging using eyes-open (EO) EEG is challenging due to stimulus-driven nonstationarity and fragmented oscillatory responses. This study aims to determine whether prototype-based representations derived from Dynamic Mode Decomposition (DMD) can improve AD detection from EO photostimulation EEG. Methods: We propose a DMD-based framework termed DMD-based Clustered Pattern Projection (DMD-CPP). Segment-wise DMD representations were clustered to learn class-specific medoid prototypes, and each EEG epoch was encoded as a vector of cosine-similarity coordinates with respect to these prototypes. A linear SVM classifier was trained on the resulting DMD-CPP features and evaluated under strict leave-one-subject-out validation. Results: The DMD-CPP model achieved competitive classification accuracy and, importantly, enhanced margin-based reliability. In EO photostimulation, AD versus healthy control classification showed a pronounced improvement, characterized by wider and more asymmetric decision margins, particularly assigning low confidence to normal epochs misclassified as AD. Tasks involving frontotemporal dementia also showed improvement, although the effect was less pronounced than for AD. Conclusions: Clustering-based pattern projection has been shown to stabilize EEG dynamics and provide an interpretable, confidence-aware feature representation. These findings suggest that DMD-CPP offers a promising framework for reliable AD detection from EO EEG, where conventional spectral methods typically struggle.