Event Synchronization in the Construction of an EEG Motor Execution Database

Abstract

Electroencephalography (EEG) based brain‑computer interfaces rely on well-labeled datasets to train and validate algorithms. However, available EEG datasets forego the synchronizing methodology, lowering reproducibility and compromising label accuracy. This work describes a photodiode‑based synchronization strategy, combined with a vertical‑synchronization stimulus presentation to accurately timestamp motor execution events during data collection.