Camera-based eye tracking improves the signal-to-noise ratio of EEG
Bibliographic record
Abstract
To examine event-related potentials (ERPs) - the brain responses associated with specific sensory, cognitive, and motor processes - researchers record brain activity using electroencephalography (EEG) while participants perform carefully designed tasks. EEG is inherently noisy, hence ERPs are derived by averaging over many trials. Since eye movements and blinks generate large electrical signals that contaminate EEG, methods have been developed to deal with these artifacts. Electrooculography uses activity from electrodes near the eyes to identify and remove contaminated trials, or participants who move their eyes too often. This method reduces reduces the signal-to-noise ratio and/or increases the number of participants required. Mathematical techniques, such as independent component analysis, have been used to allegedly remove eye movement activity. However, it is unclear whether all such activity is definitively removed, while leaving the neural activity of interest unaltered. Moreover, this approach necessarily includes trials with eye movements, often contrary to task requirements. In four recent experiments combining camera-based eye tracking and EEG, we demonstrated the potential of combining these technologies to increase ERP data quality. We used a cueing task where participants maintained fixation, ignored or made eye movements to uninformative exogenous or endogenous cues, then made manual localization responses to targets. Eye tracking ensured appropriate oculomotor behavior at all times. Error messages were presented when incorrect eye movements were observed and such trials were recycled. The ERPs were much cleaner (enhanced signal-to-noise ratio) here, relative to similar experiments without concurrent eye tracking. There are at least two (not mutually exclusive) explanations for this observation: a) participants learn to learn to control their oculomotor behavior through online feedback, minimizing untoward blinks and eye movements, and b) eye monitoring allows more accurate categorization of trials to be excluded than traditional techniques. Meeting abstract presented at VSS 2013
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.007 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.004 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".