Invited Symposium 3 - 100 Years of EEG: Where Are We?

Invited Symposium 3: Tuesday, April 1, 2025, 10:00 am – 12:00 pm EDT, Grand Ballroom

Chair: Bin He¹; ¹Carnegie Mellon University
Presenters: Bin He, Christoph Michel, Laura Astolfi, Scott Makeig

Electroencephalography (EEG) represents the scalp manifestation of brain activity, reflecting synchronized activity of neuronal populations engaged in specific tasks or brain states. Over the past century, EEG research has undergone remarkable evolution. Initially focused on the visual inspection of amplitude and frequency changes in spontaneous brain activity—such as rhythmic oscillations in the delta, theta, alpha, beta, and gamma bands—and event-related potentials (ERPs), it has advanced to enable three-dimensional source imaging of dynamic brain activity, microstate analysis of brain information processing, separation of brain signals from non-brain components, and the study of coherence and functional connectivity. In this session, Dr. Bin He will discuss EEG source imaging for dynamic brain source localization using high-density EEG, as well as EEG-based brain-computer interfaces for controlling computers and robotic devices. Dr. Christoph Michel will discuss EEG microstate analysis and its role in revealing information processing within large-scale neural networks. Dr. Laura Astolfi will discuss the investigation of functional brain networks through analyses of synchrony, coherence, and causality. Dr. Scott Makeig will discuss EEG information mining using independent component analysis (ICA) to extract brain-relevant sources and reject non-brain processes.

Presentations

EEG Source Imaging and Brain-Computer Interface

Bin He¹; ¹Carnegie Mellon University

Brain activity is distributed across a three-dimensional volume and evolves over time. Mapping the spatiotemporal distribution of brain activation with high spatial and temporal resolution is crucial for understanding brain function and aiding in the clinical diagnosis and management of brain disorders. EEG source imaging has significantly advanced our ability to map and image brain function and dysfunction. In this presentation, we will review the EEG source imaging approaches developed over the past decades and highlight the state-of-the-art capabilities for imaging brain source location, extent, and dynamics using scalp-recorded EEG as a functional neuroimaging modality. We will also discuss the principles and advancements in brain-computer interfaces utilizing noninvasive EEG, which enable the decoding of human intentions for controlling computers and robots. We demonstrate that humans can control the flight of a drone and robotic arm—enabling actions such as reaching, grasping, and continuous movement in three-dimensional space—using only "thoughts" decoded from noninvasive EEG signals.

Temporal Dynamics of Evoked and Spontaneous Neuronal Networks Revealed by Microstate Analysis of High-density EEG

Christoph Michel¹; ¹Faculty of Medicine, University of Geneva, Switzerland

High-density EEG recordings, along with scalp electric field analysis, provide a powerful tool for capturing the rapid flow of information processing within large-scale neural networks. EEG studies examining the spatial distribution of global scalp electric fields have demonstrated that both ERPs and ongoing EEG activity can be segmented into brief, sub-second periods of stable topographies, referred to as "EEG microstates." These microstates are hypothesized to represent fundamental units of thought during information processing. Neurophysiologically, EEG microstates correspond to periods of synchronized activity within large-scale networks that support specific cognitive functions. EEG microstate analysis has become widely adopted in EEG research, with a notable increase in publications across cognitive and clinical neuroscience. This presentation provides an overview of the analytical approach and summarizes current knowledge on the functional significance of EEG microstates.

Exploring Functional Brain Networks with EEG: Synchrony, Coherence, and Causality

Laura Astolfi¹; ¹University of Rome Sapienza, Italy

Over its century-long history, EEG analysis has evolved from visual inspection of amplitude and frequency changes over time to a comprehensive description of data properties in the temporal, spatial and spectral domains. Unlike other neuroimaging techniques, EEG allows the study of fast neural interactions, which are essential for understanding synchrony, coherence and causal relationships within functional brain networks. Despite limitations in spatial resolution, the millisecond-level precision of EEG is essential for capturing the dynamics of brain connectivity, allowing for a deeper analysis of how neuronal populations communicate and synchronise over time. My presentation will explore these aspects, highlighting how EEG data can reveal the temporal flow and structure of neural interactions in cognitive and clinical contexts. By situating EEG-based connectivity studies within the larger field of network neuroscience, I will discuss current methodologies, recent advances and ongoing challenges in using EEG to reveal the brain's dynamic network organization in health and disease.

EEG Information Mining

Scott Makeig¹; ¹University of California San Diego

The primary limitation of EEG as a brain imaging modality, the broad spatial spread of current from each cortical location to the scalp, is ever more addressable through information theory-based methods. Separation of potentials arising from brain and non-brain processes by independent component analysis (ICA) is now routine. Identification of multiple temporally and functionally-distinct brain ‘effective source’ signals is an important concomitant. Localization of cortical brain effective source activities contributing to the scalp-recorded signals is most accurate when combined with geometric information in participant magnetic resonance head images. Combining these with ICA also enables estimation of individual participant skull conductivity – the strongest uncertainty now limiting localization of specific cortical territories contributing most markedly to scalp-recorded signals. ICA decomposition of scalp EEG recordings can now also be used to detect non-stationary transitions in brain and cognitive state, to identify characteristic source frequency modes, and to clarify relationships of brain source activities to individual event context.