Estimating the Event-Related Potential from Few EEG Trials

TL;DR

This paper presents EEG2ERP, a deep learning autoencoder that estimates event-related potentials from few EEG trials, outperforming traditional averaging methods and enabling ERP analysis with fewer trials across diverse datasets.

Contribution

Introduction of EEG2ERP, a novel uncertainty-aware autoencoder that accurately estimates ERPs from limited EEG trials, including a variance decoder for uncertainty modeling.

Findings

EEG2ERP outperforms conventional averaging in few-trial scenarios.

Method generalizes well across multiple datasets and subjects.

First deep learning approach to map EEG to ERP.

Abstract

Event-related potentials (ERP) are measurements of brain activity with wide applications in basic and clinical neuroscience, that are typically estimated using the average of many trials of electroencephalography signals (EEG) to sufficiently reduce noise and signal variability. We introduce EEG2ERP, a novel uncertainty-aware autoencoder approach that maps an arbitrary number of EEG trials to their associated ERP. To account for the ERP uncertainty we use bootstrapped training targets and introduce a separate variance decoder to model the uncertainty of the estimated ERP. We evaluate our approach in the challenging zero-shot scenario of generalizing to new subjects considering three different publicly available data sources; i) the comprehensive ERP CORE dataset that includes over 50,000 EEG trials across six ERP paradigms from 40 subjects, ii) the large P300 Speller BCI dataset, and iii) a neuroimaging dataset on face perception consisting of both EEG and magnetoencephalography (MEG) data. We consistently find that our method in the few trial regime provides substantially better ERP estimates than commonly used conventional and robust averaging procedures. EEG2ERP is the first deep learning approach to map EEG signals to their associated ERP, moving toward reducing the number of trials necessary for ERP research. Code is available at https://github.com/andersxa/EEG2ERP