Preliminary Results on a New Algorithm for Blink Correction Adaptive to Inter- and Intra-Subject Variability

TL;DR

This paper introduces a new EEG preprocessing algorithm that automatically corrects blinking artifacts without extra electrodes, improving brain-computer interface accuracy by adapting to individual and session variability.

Contribution

The novel ABC algorithm automatically adapts to blink variability in EEG signals, eliminating the need for additional EOG electrodes and enhancing classification accuracy.

Findings

13.7% mean accuracy increase with ABC

Effective correction of blink artifacts in EEG

No additional EOG electrodes required

Abstract

This paper presents a new preprocessing method to correct blinking artifacts in Electroencephalography (EEG) based Brain-Computer Interfaces (BCIs). This Algorithm for Blink Correction (ABC) directly corrects the signal in the time domain without the need for additional Electrooculogram (EOG) electrodes. The main idea is to automatically adapt to the blink's inter- and intra-subject variability by considering the blink's amplitude as a parameter. A simple Minimum Distance to Riemannian Mean (MDRM) is applied as the classification algorithm. Preliminary results on three subjects show a mean classification accuracy increase of 13.7% using ABC.