The Dynamics of Error Processing in the Human Brain as Reflected by High-Gamma Activity in Noninvasive and Intracranial EEG

Martin A. Volker, Lukas D. J. Fiederer, Sofie Berberich, Jiří Hammer, Joos Behncke, Pavel Kršek, Martin Tomášek, Petr Marusič, Peter C. Reinacher, Volker A. Coenen, Moritz Helias, Andreas Schulze‐Bonhage, Wolfram Burgard, Tonio BallbioRxiv 2017 · DOI

Abstract

Error detection is a fundamental cognitive function, yet the underlying neural dynamics have remained partially obscured. While traditional low-frequency EEG components like the Error-Related Negativity (ERN) provide a snapshot of this process, they lack the spatial and temporal resolution to capture local cortical processing. This study utilizes high-gamma band (HGB) activity (>50 Hz) to map error processing in the human brain. Using a combination of optimized noninvasive scalp EEG in 31 healthy subjects and intracranial EEG (iEEG) in 9 epilepsy patients, researchers identified a sustained "cascade" of neural activity. This activity begins in anterior regions and migrates to the parietal cortex, persisting for over a second. These findings establish HGB as a robust neural correlate of error processing, revealing a network that extends to deep structures like the hippocampus and insula.

Introduction: Beyond the "Blurry" Signal

When we make a mistake, our brains react with remarkable speed. For decades, cognitive neuroscientists have tracked this reaction using Error-Related Potentials (ERPs) in low-frequency EEG. Specifically, the Error-Related Negativity (ERN) and Error Positivity (Pe) have served as the primary markers for error detection. However, these signals are spatially diffuse and do not fully reflect the high-frequency, local cortical computations that characterize complex information processing.

High-Gamma Band (HGB) activity (50–150 Hz) offers a much clearer view of these local dynamics. The challenge, however, is that HGB signals are extremely faint and easily masked by noise from muscles or eye movements when measured from the scalp. This study sought to overcome these technical hurdles to determine if HGB can reliably track error processing noninvasively and what it reveals about the brain's "mistake-correction" network.

Methods: A Dual-Modality Approach

To capture these elusive signals, the researchers employed two complementary methods:

  1. Optimized Noninvasive EEG: 31 healthy participants were monitored using a high-resolution 128-channel EEG setup within an electromagnetically shielded environment. To ensure HGB signals were neural and not artifactual, the team used high-precision eye tracking to account for microscopic eye movements (microsaccades).
  2. Intracranial EEG (iEEG): To validate the scalp findings, data were collected from 9 patients with implanted depth electrodes. This provided a "gold-standard" view of local cortical activity.

Participants performed the Eriksen flanker task, which requires identifying a central target while ignoring distracting "flanker" stimuli. This task is designed to induce occasional fast, impulsive errors.

Experimental paradigm using the Eriksen flanker task with audiovisual feedback.

As shown in Figure 1, the task involved a sequence of fixation, cue, and stimulus, followed by a response window. Participants received audiovisual feedback: a happy face for correct responses, a sad face for errors, and a snail icon for slow responses.

Key Findings: The Long Tail of Error Processing

The study's results challenge the notion that error processing is a fleeting, momentary event.

The Frontal-to-Parietal Shift

The HGB response followed a distinct spatio-temporal path. It began in the anterior (frontal) regions, peaking approximately 150–200 ms after the error. Subsequently, the activity shifted to the medial parietal regions, peaking between 600 and 700 ms.

Sustained Neural "Rumination"

One of the most striking findings was the duration of the HGB signal. While traditional ERPs typically dissipate within a few hundred milliseconds, HGB activity in the parietal cortex remained significantly elevated for up to 1.5 seconds post-error. This suggests that the brain continues to process the mistake long after the initial detection, likely engaging in error-related learning or behavioral adjustment.

Deep Network Involvement

The intracranial data revealed that the error-processing network is far more extensive than previously thought. Beyond the expected frontal and parietal areas, significant HGB increases were observed in the insula and the hippocampus. The involvement of the hippocampus is particularly noteworthy, as it suggests a direct link between the immediate detection of a mistake and the formation of memories to prevent its recurrence.

Technical Validation and Significance

A major contribution of this work is the technical validation of noninvasive HGB mapping. The researchers demonstrated that the spectral patterns found in scalp EEG closely matched those in direct cortical recordings, particularly in the 60–90 Hz range. Crucially, they proved that these signals remained even after controlling for ocular artifacts, debunking the common criticism that scalp HGB is merely "eye noise."

Furthermore, the study found that HGB amplitude did not correlate with the amplitude of traditional ERN/Pe signals. This indicates that HGB provides a unique, independent functional window into the brain's dynamics that traditional methods simply miss.

Conclusion and Future Directions

By mapping the high-frequency dynamics of error processing, this research provides a more comprehensive picture of how the human brain monitors its own behavior. The discovery of a sustained, multi-region cascade of activity opens new avenues for studying cognitive control.

Future applications of this HGB mapping could include:

  • Clinical Diagnostics: Investigating altered error processing in psychiatric conditions like OCD or depression.
  • Brain-Machine Interfaces (BMIs): Developing systems that can detect a user's error in real-time and automatically "undo" the action.
  • Learning Science: Understanding how the "long tail" of error processing contributes to long-term skill acquisition and behavioral adaptation.

Summary updated

Abstract

Abstract Error detection in motor behavior is a fundamental cognitive function heavily relying on cortical information processing. Neural activity in the high-gamma frequency band (HGB) closely reflects such local cortical processing, but little is known about its role in error processing, particularly in the healthy human brain. Here we characterize the error-related response of the human brain based on data obtained with noninvasive EEG optimized for HGB mapping in 31 healthy subjects (15 females, 16 males), and additional intracranial EEG data from 9 epilepsy patients (4 females, 5 males). Our findings reveal a comprehensive picture of the global and local dynamics of error-related HGB activity in the human brain. On the global level as reflected in the noninvasive EEG, the error-related response started with an early component dominated by anterior brain regions, followed by a shift to parietal regions, and a subsequent phase characterized by sustained parietal HGB activity. This phase lasted for more than 1 s after the error onset. On the local level reflected in the intracranial EEG, a cascade of both transient and sustained error-related responses involved an even more extended network, spanning beyond frontal and parietal regions to the insula and the hippocampus. HGB mapping appeared especially well suited to investigate late, sustained components of the error response, possibly linked to downstream functional stages such as error-related learning and behavioral adaptation. Our findings establish the basic spatio-temporal properties of HGB activity as a neural correlate of error processing, complementing traditional error-related potential studies. Significance Statement There is great interest to understand how the human brain reacts to errors in goal-directed behavior. An important index of cortical and subcortical information processing is fast oscillatory brain activity, particularly in the high-gamma band (above 50 Hz). Here we show that it is possible to detect signatures of errors in event-related high-gamma responses with noninvasive techniques, characterize these responses comprehensively, and validate the EEG procedure for the detection of such signals. In addition, we demonstrate the added value of intracranial recordings pinpointing the fine-grained spatio-temporal patterns in error-related brain networks. We anticipate that the optimized noninvasive EEG techniques as described here will be helpful in many areas of cognitive neuroscience where fast oscillatory brain activity is of interest.
BibTeX
@misc{volker2017dynamics,
  author = {Martin A. Volker and Lukas D. J. Fiederer and Sofie Berberich and Jiří Hammer and Joos Behncke and Pavel Kršek and Martin Tomášek and Petr Marusič and Peter C. Reinacher and Volker A. Coenen and Moritz Helias and Andreas Schulze‐Bonhage and Wolfram Burgard and Tonio Ball},
  title = {{The Dynamics of Error Processing in the Human Brain as Reflected by High-Gamma Activity in Noninvasive and Intracranial EEG}},
  year = {2017},
  doi = {10.1101/166280},
  howpublished = {bioRxiv},
}
Figure