Recurrent Neural Networks for P300-based BCI

TL;DR

This paper evaluates recurrent neural networks for detecting P300 signals in EEG data from RSVP protocols, comparing their performance to LDA and CNN, and finds RNNs with CNN components excel in transfer learning and noise resilience.

Contribution

It introduces the use of RNNs for P300 detection in RSVP data and compares their effectiveness to traditional and CNN-based methods, highlighting advantages in transfer learning and noise robustness.

Findings

LDA performs as well or better than neural networks on single subjects.

CNN-RNN hybrid models improve transfer learning across subjects.

RNN-based models are more resilient to temporal noise.

Abstract

P300-based spellers are one of the main methods for EEG-based brain-computer interface, and the detection of the P300 target event with high accuracy is an important prerequisite. The rapid serial visual presentation (RSVP) protocol is of high interest because it can be used by patients who have lost control over their eyes. In this study we wish to explore the suitability of recurrent neural networks (RNNs) as a machine learning method for identifying the P300 signal in RSVP data. We systematically compare RNN with alternative methods such as linear discriminant analysis (LDA) and convolutional neural network (CNN). Our results indicate that LDA performs as well as the neural network models or better on single subject data, but a network combining CNN and RNN has advantages when transferring learning among subejcts, and is significantly more resilient to temporal noise than other methods.