Work

Interpretable feature-transformer framework for cross-subject MCI detection using nonlinear dynamical and graph-theoretic EEG features