Towards Effective Deep Neural Network Approach for Multi-Trial P300-based Character Recognition in Brain-Computer Interfaces

TL;DR

This paper introduces a novel weighted ensemble spatio-sequential CNN to enhance P300 detection and character recognition in brain-computer interfaces, addressing challenges like low SNR and uneven P300 distribution.

Contribution

The study proposes a new deep learning model, WE-SPSQ-CNN, that improves classification accuracy and robustness for P300-based BCI character recognition tasks.

Findings

Achieved 69.7% and 79.9% P300 classification accuracy for two subjects.

Attained 76.5%, 87.5%, and 94.5% character recognition accuracy with increasing repetitions.

Outperformed existing models in five-repetition scenarios.

Abstract

Brain-computer interfaces (BCIs) enable direct interaction between users and computers by decoding brain signals. This study addresses the challenges of detecting P300 event-related potentials in electroencephalograms (EEGs) and integrating these P300 responses for character spelling, particularly within oddball paradigms characterized by uneven P300 distribution, low target probability, and poor signal-to-noise ratio (SNR). This work proposes a weighted ensemble spatio-sequential convolutional neural network (WE-SPSQ-CNN) to improve classification accuracy and SNR by mitigating signal variability for character identification. We evaluated the proposed WE-SPSQ-CNN on dataset II from the BCI Competition III, achieving P300 classification accuracies of 69.7\% for subject A and 79.9\% for subject B across fifteen epochs. For character recognition, the model achieved average accuracies of 76.5\%, 87.5\%, and 94.5\% with five, ten, and fifteen repetitions, respectively. Our proposed model outperformed state-of-the-art models in the five-repetition and delivered comparable performance in the ten and fifteen repetitions.