High-gamma and beta bursts in the Left Supramarginal Gyrus can accurately differentiate verbal memory states and performance.

Cross-validated convolutional neural networks (CNNs) utilizing intracranial EEG (iEEG) recorded from the left supramarginal gyrus can label words remembered or forgotten in verbal memory encoding.The cross-validated CNN performs best when trained and tested using iEEG with higher amplitude high-gamma (80-150 Hz) and beta (16-40 Hz) signals.Cross-validated CNNs utilizing only high-gamma and beta bursts can also differentiate encoding and recall epochs and label good recall epochs from poor recall epochs.