METER: Matched-Event Calibration for Short-Event EEG Microstate Transition Analysis
Organizations: eBRAIN Lab, New York University Abu Dhabi, Abu Dhabi, United Arab Emirates. · Dubai College, Dubai, United Arab Emirates.
Abstract
Short cognitive events contain unequal and often sparse transition evidence, so raw EEG microstate distances can reflect finite sampling rather than neural reconfiguration. We introduce METER (Matched-Event Transition Estimation against a Reference), a participant-relative framework that separates marginal occupancy from conditional next-state choice and evaluates each event against duration-conditioned events from a fixed participant-specific reference bank. METER-Jump integrates uncertainty in sparse target and reference transition tables, returning a randomized event rank for testing and a continuous standardized coordinate for association analysis. Using a reproducible three-state representation selected by spatial criteria, we evaluated METER with participant-fitted simulations, estimator ablations, map perturbations, and preprocessing sensitivities. Event-level false-positive rates were 0.044-0.056 across four heterogeneous null mechanisms. By contrast, global cutoffs learned under the participant-fitted null and transferred unchanged to other regimes produced false-positive rates from 0.004 under rare transitions to 0.234 under map perturbation; duration-conditioned ranks remained near 0.05. Detection probability was approximately 0.69 with fixed maps but fell to 0.22 after bootstrap-derived map-assignment perturbation. In 4,175 natural-reading events from 18 participants, no linguistic association survived multiplicity correction in the primary 30-ms analysis, although a 20-ms sensitivity changed one within-stream max- decision. METER therefore calibrates finite-event departures conditional on a defined microstate sequence, while map assignment, filtering, temporal smoothing, and reference construction remain separate sources of measurement uncertainty.
Figures & tables
| Component | Study choice | Scientific role |
|---|---|---|
| State representation | ; 25 starts; 30-ms primary smoothing | Reproducible task-derived coordinates; 20- and 40-ms streams test temporal sensitivity. |
| Reference bank | 96 events per participant, jointly balanced over block, question status, acquisition order, and duration | Defines a participant-specific, target-disjoint baseline. |
| Jump estimator | ; 512 target draws; 128 pseudo-event draws | Integrates finite target/reference row uncertainty; concentration evaluated by simulation. |
| Confidence rule | Participant-bank 10th percentile of two-endpoint assignment margin | Omits the least certain jump boundaries while holding the rule fixed for targets. |
| Event calibration | Symmetric 95-event subreference; duration quartile; at most 50 pseudo-events | Gives real and pseudo-targets equal reference size and conditions on opportunity to observe transitions. |
| Weak-information flag | Fewer than two retained outgoing jumps in any source row | Identifies rows unable to represent both off-diagonal destinations empirically for . |
| Quantity | Median | IQR | 5th–95th percentile |
|---|---|---|---|
| Event duration (s) | 5.02 | 3.36–7.18 | 1.86–12.86 |
| EEG samples | 2,509 | 1,679–3,592 | 932–6,431 |
| Retained jumps | 46 | 29–68 | 14–126 |
| Retained jumps/s | 9.89 | 8.41–10.91 | 4.27–12.37 |
| Reference samples | 267,851 | 234,759–342,524 | 160,358–371,493 |
| Reference runs | 5,754 | 4,490–7,158 | 3,486–8,141 |
| Question | Recommendation |
|---|---|
| Are group maps stable? | Bootstrap map correspondence and quantify sequence uncertainty; map correlation alone does not establish reliable jumps. |
| Are events information-limited? | Report realized transition evidence and calibrate within participant using an opportunity variable defined independently of the transition outcome, such as duration. |
| How large should the reference bank be? | Examine a reference-size curve under the intended state alphabet and event distribution; in this study, 80–96 events was the lowest evaluated range near the fixed-map detection plateau. |
| Is the reference participant-specific? | Keep targets out of the fixed bank and repeat balanced-bank construction when effect magnitude matters. |
| Is switching reconfiguration claimed? | Compare against occupancy and distinguish first-order destination choice from higher-order sequence organization. |
| Are preprocessing or smoothing changed? | Revalidate jump counts, score agreement, event-null error, and application estimates. |
Electroencephalography
Electroencephalography (EEG) studies brain-generated electrical activity to understand cognition and support diagnosis, monitoring, and brain-computer interfaces.
See all 200 papers