Transformer Convolutional Neural Networks for Automated Artifact Detection in Scalp EEG

TL;DR

This paper introduces a CNN-transformer hybrid model with belief matching loss for automated detection of five EEG artifact types, achieving high accuracy and effective artifact rejection in multi-channel EEG analysis.

Contribution

It presents a novel CNN-transformer architecture with belief matching loss for multi-class EEG artifact detection, improving accuracy and artifact rejection capabilities.

Findings

Segment-level classifiers achieve BAC up to 0.947 for chewing artifacts.

Combined detectors reach sensitivity of 60.4% and specificity of 95% for artifact detection.

The method effectively rejects artifacts while preserving background EEG.

Abstract

It is well known that electroencephalograms (EEGs) often contain artifacts due to muscle activity, eye blinks, and various other causes. Detecting such artifacts is an essential first step toward a correct interpretation of EEGs. Although much effort has been devoted to semi-automated and automated artifact detection in EEG, the problem of artifact detection remains challenging. In this paper, we propose a convolutional neural network (CNN) enhanced by transformers using belief matching (BM) loss for automated detection of five types of artifacts: chewing, electrode pop, eye movement, muscle, and shiver. Specifically, we apply these five detectors at individual EEG channels to distinguish artifacts from background EEG. Next, for each of these five types of artifacts, we combine the output of these channel-wise detectors to detect artifacts in multi-channel EEG segments. These segment-level classifiers can detect specific artifacts with a balanced accuracy (BAC) of 0.947, 0.735, 0.826, 0.857, and 0.655 for chewing, electrode pop, eye movement, muscle, and shiver artifacts, respectively. Finally, we combine the outputs of the five segment-level detectors to perform a combined binary classification (any artifact vs. background). The resulting detector achieves a sensitivity (SEN) of 60.4%, 51.8%, and 35.5%, at a specificity (SPE) of 95%, 97%, and 99%, respectively. This artifact detection module can reject artifact segments while only removing a small fraction of the background EEG, leading to a cleaner EEG for further analysis.