Editorial: Combined EEG in research and diagnostics: Novel perspectives and improvements

Camillo Porcaro, Kamran Avanaki, Óscar Arias-Carrión, Morten MørupFrontiers in Neuroscience 2023 · DOI · OpenAlex

Introduction

The field of neuroimaging is undergoing a paradigm shift. While Electroencephalography (EEG) has long been a staple of neuroscience due to its exceptional temporal resolution, its inherent limitations in spatial localization have historically constrained its utility. This editorial, recently published in Frontiers in Neuroscience, synthesizes a collection of 11 pioneering articles that advocate for "multimodal integration"—the strategic pairing of EEG with other neuroimaging and brain stimulation modalities to create a more holistic view of brain dynamics.

The Spatial-Temporal Resolution Trade-off

A fundamental challenge in neuroimaging is the inverse relationship between spatial and temporal precision. EEG captures electrical activity at the millisecond scale, making it ideal for tracking the rapid "when" of neural processes. However, because signals are recorded at the scalp, determining the exact "where" (the source) remains difficult. Conversely, Functional Magnetic Resonance Imaging (fMRI) provides millimeter-level spatial maps but relies on blood-oxygen-level-dependent (BOLD) signals, which are significantly slower.

By integrating these technologies, researchers can bridge this gap. The articles within this Research Topic demonstrate that combining EEG with fMRI, Magnetoencephalography (MEG), or functional Near-Infrared Spectroscopy (fNIRS) allows for a comprehensive mapping of brain function that neither tool could achieve in isolation.

Methodological Innovations in Multimodal Synergy

The integration of disparate data streams requires sophisticated mathematical and technical frameworks. Several key methodologies highlighted in this collection include:

  • C-SECSI (Coupled Semi-Algebraic framework): This tensor decomposition method is designed to separate meaningful brain oscillations from background noise. It is particularly effective in MEG-EEG fusion, where it helps extract physiologically relevant signals from high-dimensional data.
  • Microstate Modeling: By analyzing the topographic transitions of brain activity, researchers can identify subject-specific archetypes that represent the fundamental building blocks of cognition.
  • Graph Theory and Network Topology: Using "weighted visibility graphs," researchers can extract topological features from EEG channels, providing insights into the functional connectivity of the brain.
  • Deep Learning: The application of computer vision models to EEG spectral and topographic images has shown promise in automating the differentiation between healthy aging and neurodegenerative conditions.

Clinical Applications and Biomarkers

The ultimate goal of these multimodal approaches is to improve clinical outcomes. The research presented spans a wide range of conditions, including Alzheimer’s Disease (AD), Parkinson’s, Stroke, Depression, and Epilepsy.

One of the most significant insights in this collection is the realization that EEG features—such as network topology or microstate changes—can serve as objective biomarkers. For instance, deep learning models trained on EEG topographic maps have successfully identified early signs of neurodegeneration. Furthermore, in stroke rehabilitation, monitoring the brain's real-time response to therapy through combined EEG-TMS (Transcranial Magnetic Stimulation) setups offers a path toward personalized recovery protocols.

The Future: Closed-Loop Systems and Technical Challenges

A recurring theme in the editorial is the move toward "closed-loop" brain stimulation. Rather than applying static electrical currents (tES), future systems will use real-time neuroimaging to monitor the brain's state and adjust stimulation parameters dynamically. This ensures that the intervention is precisely timed to the patient's current neural activity.

However, significant hurdles remain. The "artifact problem"—the massive electrical interference generated when recording EEG during simultaneous stimulation or inside an fMRI scanner—continues to be a major technical barrier. Achieving fully continuous, clean recordings is essential for the transition from research labs to daily clinical practice.

Conclusion

The integration of EEG with complementary technologies is no longer just an experimental curiosity; it is becoming the standard for high-fidelity neuroimaging. While technical challenges regarding signal de-noising and clinical validation persist, the synergy between these modalities offers unprecedented insights into the complex dynamics of the human brain. As these tools become more refined, they will undoubtedly lead to more precise diagnostics and more effective, personalized treatments for neurological and psychiatric disorders.

This work was partly supported by the Department of Excellence 2018-2022 initiative of the Italian Ministry of Education (MIUR) awarded to the Department of Neuroscience-University of Padua [MART_ECCELLENZA18_01].

Summary updated

Abstract

sponsorship: This work was partly supported by the Department of excellence 2018-2022 initiative of the Italian Ministry of Education (MIUR) awarded to the Department of Neuroscience-University of Padua [MART_ECCELLENZA18_01]. (Department of excellence 2018-2022 initiative of Italian Ministry of Education (MIUR)|MART_ECCELLENZA18_01)
BibTeX
@inproceedings{porcaro2023editorial,
  author = {Camillo Porcaro and Kamran Avanaki and Óscar Arias-Carrión and Morten Mørup},
  title = {{Editorial: Combined EEG in research and diagnostics: Novel perspectives and improvements}},
  booktitle = {Frontiers in Neuroscience},
  year = {2023},
  doi = {10.3389/fnins.2023.1152394},
}
Figure