Single-channel EEG completion using Cascade Transformer

TL;DR

This paper introduces a Cascade Transformer-based method for completing incomplete single-channel EEG signals, achieving performance comparable to multi-channel solutions and demonstrating feasibility for single-channel EEG completion.

Contribution

The paper proposes a novel Cascade Transformer architecture and loss weighting technique specifically for single-channel EEG completion, reducing error rates significantly.

Findings

Reduced NRMSE by 2.8% and 8.5% with the new method.

Achieved NRMSE between 0.026 and 0.063 for 1%-50% missing data.

Performed comparably to multi-channel EEG completion methods.

Abstract

It is easy for the electroencephalogram (EEG) signal to be incomplete due to packet loss, electrode falling off, etc. This paper proposed a Cascade Transformer architecture and a loss weighting method for the single-channel EEG completion, which reduced the Normalized Root Mean Square Error (NRMSE) by 2.8% and 8.5%, respectively. With the percentage of the missing points ranging from 1% to 50%, the proposed method achieved a NRMSE from 0.026 to 0.063, which aligned with the state-of-the-art multi-channel completion solution. The proposed work shows it's feasible to perform the EEG completion with only single-channel EEG.