A Tale of Single-channel Electroencephalogram: Devices, Datasets, Signal Processing, Applications, and Future Directions

TL;DR

This paper presents a comprehensive narrative review of single-channel EEG, covering devices, datasets, signal processing, applications, and future directions. It outlines definitions of bipolar and unipolar configurations, surveys hardware trends and wearable devices, and catalogs public datasets, highlighting scarcity and standardization issues. It reviews signal processing pipelines, from preprocessing and artifact removal to feature extraction/selection and ML/DL modeling, emphasizing AI-based approaches and on-device processing. It surveys wide-ranging applications in sleep staging, emotion recognition, neurofeedback, education, depression, dementia (AD/MCI), epilepsy, and other domains, and discusses future directions including AI-generated data, transfer learning, edge computing, hardware innovations, and IoT integration. The findings suggest single-channel EEG, aided by advanced processing and hardware, can approach or match multichannel performance in many tasks while enabling affordable, long-term brain monitoring.

Abstract

Single-channel electroencephalogram (EEG) is a cost-effective, comfortable, and non-invasive method for monitoring brain activity, widely adopted by researchers, consumers, and clinicians. The increasing number and proportion of articles on single-channel EEG underscore its growing potential. This paper provides a comprehensive review of single-channel EEG, focusing on development trends, devices, datasets, signal processing methods, recent applications, and future directions. Definitions of bipolar and unipolar configurations in single-channel EEG are clarified to guide future advancements. Applications mainly span sleep staging, emotion recognition, educational research, and clinical diagnosis. Ongoing advancements of single-channel EEG in AI-based EEG generation techniques suggest potential parity or superiority over multichannel EEG performance.

Figures (7)

• Figure 1: Scalp electrode placement in 10–20 system for EEG acquisition 10-20.
• Figure 2: Illustration of bipolar/unipolar single-channel EEG polar.
• Figure 3: Total number of general and single-channel EEG publications over the past 5 years, as searched on Google Scholar. Single-channel EEG is an increasingly popular research topic.
• Figure 4: Flowchart of single-channel EEG signal processing illustrating key steps from signal acquisition to downstream tasks, encompassing preprocessing, feature extraction, feature selection, as well as traditional machine learning and deep learning methods.
• Figure 5: The Arousal–Valence Space (AVS) Model used in psychology and affective computing to represent and measure human emotions AVS.