- Rating
- 3(+1)
- Confidence
- 2
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
This paper proposes a multi-scale discrete spatiotemporal framework named NeuroDETR, designed to address the problem of asynchronous event detection in continuous EEG signals. Unlike traditional methods relying on sliding-window classification or fixed-segment analysis, this approach employs a DETR-like set prediction paradigm, enabling the joint prediction of event categories, onset times, and offset times directly from continuous, unsegmented EEG recordings.
Critical strengths
- This paper focuses on asynchronous event detection in continuous EEG signals, enabling the joint identification of event categories and the localization of temporal boundaries; compared to methods relying on fixed-window classification, this approach better aligns with the detection requirements of continuous brain-computer interface applications.
- The paper conducts experiments on eight datasets covering a variety of EEG tasks, as well as differing event durations and frequencies, thereby facilitating an assessment of the method's applicability from multiple perspectives.
- In addition to comparisons with various temporal detection methods and DANCE, the authors examined the roles of quantization mechanisms, time scales, and spatial hierarchies, and analyzed event localization performance across different temporal IoU thresholds, providing preliminary evidence to understand the model's performance advantages.
Critical weaknesses
- NeuroDETR draws upon various components—including VQ, hierarchical spatial encoding, sparse attention, and DETR—yet the paper has not fully demonstrated the unique technical advantages of this combination over existing EEG event detection methods.
- Temporal quantization scales, spatial hierarchical partitioning, event-density-based guidance, and dual-branch fusion all involve manually designed choices, yet there is a lack of in-depth analysis regarding the rationale and applicability limits of these designs.
- Although the paper is set against the backdrop of continuous EEG monitoring, it primarily employs a windowed input approach for training and evaluation; consequently, it lacks sufficient discussion regarding the handling of event boundaries across windows, real-time inference latency, and stability in online scenarios.
Questions for authors
- What is the core original contribution of this paper—the discrete spatiotemporal representation, the event-guided mechanism, or the collaborative learning mechanism between the two? Specifically, where do the performance gains come from when compared to existing query-based EEG detection methods such as DANCE?
- The paper addresses this issue by employing probability thresholds and a Top-K minimum selection strategy; however, there is currently a lack of systematic analysis regarding this mechanism in scenarios characterized by low event prevalence, class imbalance, and weak events. In particular, while the authors attribute some performance gains to FOCAL's ability to handle low-prevalence events, existing ablation studies—which primarily compare results between the full module and the module with specific components removed—have not yet directly demonstrated a consistent improvement in recall for weak events.
- The paper does not adequately clarify whether future signals need to be buffered during deployment, how duplicate detections between adjacent windows are handled, or whether event merging rules across windows introduce additional latency. If the model requires full computation over extended time windows, then—even if offline localization performance is strong—one cannot directly infer its suitability for strict real-time scenarios.
If the author addresses the questions raised above, I will raise my score.
5 comments
Answer (2/3)
Continuing W2 from Answer (1/3):
- FOCAL Routing (): Target events are often sparse and weak against irregular background. follows high-recall proposal filtering, retaining moderate-confidence windows over a hard cutoff for local refinement, while guarantees non-empty local attention (pure-background windows remain suppressed by decoder Class 0 without forced false alarms). Limit: Under high event occupancy, routing may saturate, causing FOCAL to approach standard local–axial attention.
- Temporal-First Fusion & Gate Init (): EEG has millisecond temporal precision but coarse spatial resolution; asynchronous detection requires exact onset/offset anchoring while spatial topography aids class discrimination. Initializing prevents coarse-binned spatial updates from perturbing early boundary regression while keeping non-zero gradients. Limit: On boundary-free, purely spatial stationary tasks (e.g., resting-state classification), this may slow early spatial convergence and add temporal overhead.
W3, Q3: Thank you for raising this deployment question. For our cross-window boundary protocol and statistics (retaining events with both onset and offset fully within the window, strictly shared across all baselines at non-overlapping stride ), please see [Reviewer Woxn, W4 & Q4]. We provide a more in-depth discussion in Appendix B. Below we clarify buffering, cross-window handling, latency, and stability:
- Offline evaluation protocol: Our NeuroDETR adopts an offline evaluation protocol in which continuous EEG is partitioned at a non-overlapping stride () and each window is inferred strictly independently—detecting events in the current window never buffers or accesses subsequent windows, relying solely on bidirectional context within the current -second window. Meanwhile, because adjacent test windows are disjoint under the complete-interval evaluation criterion, duplicate detections between adjacent windows naturally do not occur offline, requiring no cross-window event-merging post-processing.
- Online streaming deployment (chunk buffering, overlap deduplication, and merging delay): When deploying this non-causal windowed model in continuous online scenarios, the operational rules are adapted as follows:
(i) Future signal buffering: Although awaiting subsequent windows is unnecessary, symmetric dilated convolutions, FOCAL axial attention, and set decoding rely on bidirectional intra-window context, so initial inference buffers the full -second window and thereafter updates incrementally at a sliding stride (introducing intra-window lookahead buffering).
(ii) Adjacent-window deduplication and merging delay: Introducing overlapping sliding windows () to capture boundary-crossing events requires center-zone retention or 1D IoU/endpoint stitching to eliminate duplicate detections in overlap regions. Because each window outputs at most sparse intervals, scalar interval matching takes only microseconds (, negligible compute latency); the sole additional wait arises when an ongoing event crossing the current window's right edge must await the next buffer step () to finalize its offset, alongside potential sensitivity of heuristic stitching to boundary confidence jitter.
- Full-window computation and strict real-time applicability boundary: We fully agree that offline localization gains cannot be directly extrapolated to zero-lookahead strict real-time control. Strict real-time responsiveness is bounded primarily by the data-acquisition buffering stride above. Accordingly, our current framework is directly suited for offline auto-annotation and trigger-free near-real-time monitoring tolerant of second-scale buffering (e.g., clinical seizure screening/alerting and passive state assessment), whereas evolving from heuristic cross-window stitching to causal masking and end-to-end streaming cross-window memory propagation is a core focus of our future work toward sub-second closed-loop control.
The efficiency discussion, stability analysis, and remaining responses continue in Answer (3/3).
Answer (3/3)
Continuing W3 and Q3 from Answer (2/3):
Measured Long-Window Latency & Strict Real-Time Scope We benchmarked all models across all eight full test sets ( EEG, warmed up) on a single NVIDIA A100 GPU (Table R1), reporting total end-to-end Runtime (h), 8-dataset mean median single-window Latency (ms) at default , and compute per second of EEG (ms/s):
Table R1: Inference efficiency across all eight test sets ( EEG on NVIDIA A100).
| Method | DOSED | EventNet | ActionFormer | DyFADet | DANCE | NeuroDETR |
|---|---|---|---|---|---|---|
| Runtime (h) | 0.16 | 0.15 | 0.16 | 0.19 | 0.33 | 0.26 |
| Latency (ms) | 0.52 | 0.90 | 3.61 | 8.23 | 36.28 | 27.52 |
| (ms/s) | 0.04 | 0.06 | 0.25 | 0.56 | 1.03 | 0.81 |
Processing time. NeuroDETR processed 226.19 h of test EEG in 0.26 h on one NVIDIA A100. Its mean of the eight dataset-specific median processing times was 27.52 ms per acquired window, compared with 36.28 ms for DANCE. These measurements exclude signal acquisition and cross-window stitching.
Stability over long recordings. We consider sustained artifacts and gradual changes in electrode impedance potential sources of instability during long recordings. NeuroDETR processes each window independently and keeps its codebooks fixed at inference. These mechanisms prevent hidden-state errors from accumulating across windows and avoid drift caused by online codebook updates, which may help limit cumulative instability over long sessions. **Appendix B outlines how streaming evaluation could assess first-alert delay, time to finalize event offsets, and false alarms per hour.
Q2. Weak events, foreground prevalence, and class imbalance
FOCAL uses a class-agnostic target indicating whether a subwindow contains a foreground event. Temporal routing selects subwindows for local attention. Unselected subwindows pass through background normalization and still participate in axial attention within the same input. Spatial pooling uses continuous weights derived from predicted event probabilities. The decoder can assign every event query to background, including when a subwindow has been selected for refinement.
We report two additional analyses in this response. M5 removes the FOCAL module and its density supervision and replaces density-weighted spatial pooling with adaptive average pooling. The comparison with the full model measures the combined effect of these changes.
For each test subject, we ranked events by their RMS contrast with the flanking background within 2 s on either side. We used this contrast as a proxy for signal strength and evaluated recall on the lowest third. A correct detection requires the correct class and temporal IoU above 0.5.
Table R2. Event recall (%) on the lowest-contrast third of events for each test subject.
| Variant | BI2014b | BNCI2015 | BNCI2020 | TUSZ |
|---|---|---|---|---|
| M5 (w/o FOCAL) | 54.45 | 68.20 | 36.14 | 11.04 |
| Full (NeuroDETR) | 56.33 | 70.39 | 41.86 | 19.63 |
The full model improves recall on this subset in all four evaluated datasets. The largest increase is on TUSZ, from 11.04% with M5 to 19.63% with the full model, a gain of 8.59 percentage points.
- Class-imbalanced detection (Table R3): On imbalanced P300 (BI2014b, BI2013a) and clinical seizures (TUSZ), we report equal-weighted foreground ():
Table R3: Foreground class-wise (%, ) on imbalanced benchmarks.
| Variant | BI2014b | BI2013a | TUSZ |
|---|---|---|---|
| M5 (w/o FOCAL) | 27.49 | 35.16 | 10.27 |
| Full (NeuroDETR) | 31.57 | 38.50 | 15.06 |
Compared with M5, the full model improves foreground macro-event F1 by 4.08 points on BI2014b, 3.34 points on BI2013a, and 4.79 points on TUSZ. This metric gives each foreground class equal weight.
Answer (1/3)
Thank you for your thoughtful review.
W1, Q1:
1. Core Original Contribution: Synergistic Event-Guided Discrete Spatiotemporal Memory Our core contribution is not a modular assembly of sparse attention, VQ, and DETR, but a co-adapted, event-guided discrete spatiotemporal memory mechanism for asynchronous EEG—where event-presence routing (FOCAL), multiscale temporal and hierarchical spatial VQ, and set-level decoding mutually reinforce one another to drive performance gains.
Unlike existing EEG tokenizers (LaBraM, NeuroLM, DeWave) that uniformly quantize pre-segmented trials for generic reconstruction, continuous asynchronous EEG is dominated by non-stationary background activity. Here, isolated components yield limited benefits; the performance leap stems directly from cross-module synergy:
2. Unique Advantages & Source of Gains over DANCE
W2: Rationale and Applicability Scope of Core Designs
Our designs reflect intrinsic EEG neurophysiology and asynchronous detection traits (non-stationary background, low SNR, unsegmented streams), using globally fixed hyperparameters grounded in physiological timescales, the 10–20 standard, and stable optimization:
The remaining component designs and deployment analysis continue in Answer (2/3).