Pith. sign in

REVIEW 4 major objections 5 minor 1 cited by

When do they StOP?: A First Step Towards Automatically Identifying Team Communication in the Operating Room

T0 review · 4 major / 5 minor · reviewed 2026-08-08 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read This paper introduces Team-OR, the first real-surgery dataset with temporal annotations of OR team briefings, and reports that its group activity detection approach outperforms existing temporal action detectors on both Time-out and StOP?

desk verdict Team-OR is a real dataset contribution; the detection 'outperforms' claim is not feature-matched and rests on 9 StOP? test positives. read the letter →

arxiv 2502.08299 v2 pith:EGPJNOHI submitted 2025-02-12 cs.CV

Keqi Chen , Lilien Schewski , Vinkle Srivastav , Joël Lavanchy , Didier Mutter , Guido Beldi , Sandra Keller , Nicolas Padoy This is my paper · ORCID
classification cs.CV
keywords temporalactiondetectiongroupactivityoperatingroomvideoanalysissurgicalteamcommunicationTime-outprotocolStOP?skeleton-basedrecognitiondataset
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Good surgical outcomes depend on how the whole operating-room team communicates, not only on the surgeon's technical skill. This paper takes the first step toward measuring those briefings automatically by introducing Team-OR, a dataset of more than one hundred hours of real laparoscopic surgeries with start and end times marked for two standardized team communication protocols: the pre-incision Time-out and the intraoperative StOP? briefing. The paper also proposes a group activity detection method that combines scene-level video features with skeleton-based individual action features and a lightweight pooling network. On this dataset the method reports the best results among existing temporal action detectors, reaching Time-out AP@0.5 of 99.45 and StOP? AP@0.5 of 20.52. The practical motivation is patient safety: automatically knowing when and whether these briefings happen could support training, compliance monitoring, and context-aware OR assistance systems.

What carries the argument

The central object is the group activity detection task: localizing the start and end times of a briefing that involves the whole OR team in an untrimmed surgery video. The pipeline extracts two complementary feature streams from a single camera view: global temporal scene features from the pretrained video model VideoMAEv2, and per-person skeleton action features obtained by detecting and tracking people, estimating their poses, and encoding the skeleton sequences with STGCN++. These are concatenated per clip and fed into a lightweight network whose two-branch max-pooling and average-pooling feature pyramid merges multi-level temporal information; a classification head predicts the activity and a regression head predicts its start and end, with Soft-NMS removing duplicate proposals and focal plus temporal DIoU losses used during training. The pooling-only design is what lets it generalize from very few positive samples.

What would settle it

Run the released pipeline separately on each of the four synchronized camera views for the same operations and compare the predicted start and end times against the multi-view ground-truth annotations; if the detections, especially for StOP?, change materially with the chosen view, the reported scores depend on an unstated camera-selection choice rather than on group activity detection as such.

Watch

Extended reading notes

Core claim

The central claim is that group communication protocols in the OR can be treated as a temporal action detection problem, provided the detector uses both global scene context and individual skeleton action features. The paper argues that existing OR datasets, which support pose estimation or scene graph generation and mostly come from simulated or short recordings, cannot support this task, and that on the new Team-OR dataset the proposed approach outperforms ActionFormer, TriDet, and TemporalMaxer on both activities. The StOP? protocol is markedly harder than Time-out, with 20.52 versus 99.45 AP@0.5, which the paper attributes to subtle visual differences between briefing and normal taskwork, imperfect protocol compliance, and very few positive examples.

Load-bearing premise

The load-bearing premise is that a single camera view is guaranteed to show all participating team members during every briefing, but the paper does not say which of the four recorded views is used or how it was chosen.

Editorial extensions

If this is right

  • The same task definition, dataset split, and evaluation protocol can serve as a baseline for future work on group activity detection in the OR.
  • The reported end-to-end throughput of about 33 frames per second suggests that such detection could run in live intra-operative support systems.
  • The large gap between Time-out and StOP? results shows that subtle, poorly compliant group activities remain a hard target even with skeleton and scene features combined.
  • The average protocol durations reported in the dataset, about 90 seconds for Time-out and 63 seconds for StOP?, give concrete targets for detector granularity and for scheduling brief monitoring windows.
  • Because the two protocols follow standardized instructions, the annotations and detection approach could be transferred to other hospitals, although the paper only evaluates one clinical site.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Going beyond the paper: whether the single camera view contains every briefing participant is the first thing to test, since the paper does not specify which view is used or how it was selected.
  • Going beyond the paper: running the released system separately on each of the four synchronized views would show how much of the reported performance depends on an unstated recording-geometry choice.
  • Going beyond the paper: the deliberate exclusion of audio leaves a strong, privacy-sensitive cue unused; a multimodal version with speech activity detection might substantially improve the StOP? results.
  • Going beyond the paper: the same sparse-event, pooling-based detector could be applied to other long videos with rare and visually subtle group activities, such as emergency team training or classroom collaboration.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper introduces Team-OR, a multi-view dataset of 37 real laparoscopic surgeries (43 videos, 105 hours) with temporal annotations of 33 Time-out and 22 StOP? team communication briefings. It proposes a group activity detection approach that concatenates VideoMAEv2 scene features with STGCN++ skeleton features and feeds them into a lightweight 1D convolutional network with max- and average-pooling pyramids, followed by classification and regression heads. On a 60/40 video-level split, the method reports AP@0.5 of 99.45 for Time-out and 20.52 for StOP?, outperforming the TAD baselines in Table 2; the paper also reports ablations and a real-time FPS figure of 33.03.

Significance. The dataset contribution is valuable and timely: Team-OR appears to be the first OR dataset of real surgeries with temporal annotations of clinically validated team communication protocols, and the authors commit to releasing code. The multi-view setup, the connection to the StOP? II trial, and the light-weight pooling design are genuine strengths. However, the central empirical claim of outperforming state-of-the-art TAD approaches is not yet established: the test set has only 9 StOP? positives, no uncertainty quantification is reported, and the comparison is not feature-matched, since the proposed model uses an extra skeleton modality. The low StOP? AP also shows that the task remains far from solved, so the paper's main value at this stage is the dataset and the task formulation rather than the detection accuracy.

major comments (4)
  1. [§4.1, Tables 1 and 2] The central performance claim rests on a test split with 13 Time-out and only 9 StOP? positive instances, and the paper reports a single run with no confidence intervals, multiple seeds, or significance tests. For StOP?, the AP@0.5 margin over TemporalMaxer is 8.14 points (20.52 vs 12.38); with 9 positives, a small change in detection ranking can plausibly shift AP by more than this margin. Please report per-seed results, confidence intervals (e.g., bootstrap over videos), or a significance test, and temper the 'outperforms' claim accordingly.
  2. [§4.4, Tables 2 and 3] The comparison with TAD baselines is not feature-matched. The proposed method uses F = (Fg, Fl), i.e., VideoMAEv2 plus STGCN++ skeleton features, while the baselines in Table 2 are evaluated with VideoMAEv2 features only. Table 3 shows that adding skeleton features is responsible for a large part of the improvement: between rows 2 and 3, Time-out AP@0.5 rises from 79.06 to 99.45 and StOP? AP@0.5 from 12.77 to 20.52. Thus the reported advantage over TemporalMaxer and TriDet may reflect the extra skeleton modality rather than the proposed pooling architecture. Please evaluate the baselines with the same concatenated features, or ablate the proposed model without skeleton features, and compare all methods on equal footing.
  3. [§3.1, Table 1] The dataset consists of 43 videos from 37 surgeries, and the train/test split is performed at the video level. Because the authors state that some operations had to be paused and therefore produced multiple videos, two videos from the same operation can in principle appear in both train and test, creating a risk of information leakage. Please clarify whether any operation contributes videos to both splits and, if so, re-split by operation or explicitly verify that no surgery spans the train/test boundary.
  4. [§3.3] The method uses 'one camera view as the input', but the paper never specifies which of the three ceiling views or the laparoscopic view is used, how that view is selected, or whether that view always contains all team members participating in the briefing. Since both protocols require all team members to participate, a view that misses a participant cannot support the claimed detection of the group activity. Please specify the view-selection procedure and discuss its coverage relative to the team layout in the OR.
minor comments (5)
  1. [§3.2] The FPR-95 experiment used to motivate the hierarchical analysis is described qualitatively, but no quantitative results are shown; please provide the actual FPR-95 values or state that they are omitted for space and available upon request.
  2. [§3.1] The annotation section mentions 'boundary errors of less than five seconds', but no inter-annotator agreement measure is reported; if available, please add one to support the reliability of the annotations.
  3. [§4.2] The implementation details include the feature stride and clip settings, but not the temporal resolution of the final proposals or the Soft-NMS threshold used at inference; please include these for reproducibility.
  4. [§4.3] The phrase 'our approach performs the best on both activities, which establishes a solid baseline' is stronger than the results justify, given the StOP? AP@0.5 of 20.52; please rephrase to acknowledge the limited test size and the remaining difficulty of the task.
  5. [Table 3] The ablation study would be more informative if each configuration were run with multiple seeds and reported with standard deviations, since the differences between some rows are small and may not be stable.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central claims rest on held-out test evaluation with externally pretrained features and external TAD baselines.

full rationale

The derivation chain is not circular. The paper's central quantitative claim — that the proposed pooling-based detector outperforms ActionFormer, TriDet, and TemporalMaxer on the Team-OR test split — is established by training on the 60% split and evaluating AP on the held-out 40% split (Tables 1-2), with no target-derived parameter used as an input. Features are extracted with externally pretrained models (VideoMAEv2, STGCN++) and compared against standard TAD baselines under the same protocol. The task definition (localizing Time-out and StOP? intervals) comes from clinical protocol literature (WHO checklist and the StOP? II trial), which is an external standard rather than an output of this paper. Ablations in Tables 3-4 show incremental contributions of the pooling branches and skeleton features; these are design components, not renamed targets. The self-citations present ([20], [21], [27], [35], [36]) are related-work, annotation-platform, or prior OR-dataset references, and the StOP? II citation [14] supplies the data-collection context; none is invoked as a uniqueness theorem or as a substitute for the empirical evaluation. Smaller concerns — only 13 Time-out and 9 StOP? test positives, single-camera input without view selection details, and no confidence intervals — bear on statistical robustness and generalizability, not on whether any result is equivalent to its inputs by construction. Section 4.5 explicitly acknowledges the limited sample and single-site evaluation. No equation in the paper defines a predicted quantity in terms of the annotation target, and no fitted parameter is relabeled as a prediction. Therefore the circularity score is 0.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central empirical claims rest on the annotation accuracy, on transferability of pretrained features to OR team communications, on the sufficiency of a single camera view, and on the adequacy of standard TAD metrics for 55 positive events. These are domain assumptions rather than mathematical axioms; none is independently verified in the paper. There are no invented entities.

free parameters (3)
  • Pooling pyramid depth L = 7
    Set by hand in Section 4.2 without sensitivity analysis; controls the number of temporal scales in Eqs. 2 and 3.
  • Feature extraction stride = 32 frames
    Chosen in Section 4.2 for clip-level features; no ablation reported for this stride.
  • Train/test split ratio = 60/40 by video
    Split by video in Section 4.1; only one split is used, so variance across splits is unknown.
assumptions (4)
  • domain assumption Temporal annotations with under five seconds of boundary error are accurate enough to serve as ground truth.
    Section 3.1 states manual boundary errors are less than five seconds but reports no inter-annotator agreement or validation against protocol logs.
  • domain assumption Pretrained VideoMAEv2 and STGCN++ features transfer to OR team communication without fine-tuning.
    Section 3.3 extracts features with pretrained models; transferability is assumed and not validated on OR-specific data.
  • domain assumption A single ceiling camera view is sufficient to observe all participating team members during briefings.
    Section 3.3 says 'we utilize one camera view as the input'; no analysis shows this view always covers the team, which is central for group activity detection.
  • domain assumption Standard temporal action detection evaluation, AP at tIoU thresholds, is appropriate for group activities with only 55 annotations.
    Section 4.1 uses standard TAD metrics; their statistical behavior on so few positives is not discussed.

how reviews work

0 comments
Cite this review

Pith. "Pith review of When do they StOP?: A First Step Towards Automatically Identifying Team Communication in the Operating Room." pith.science (2026). https://pith.science/paper/EGPJNOHI

@misc{pith2026250208299,
  author       = {Pith},
  title        = {Pith review of: When do they StOP?: A First Step Towards Automatically Identifying Team Communication in the Operating Room},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EGPJNOHI}},
  note         = {Machine review of arXiv:2502.08299}
}
read the original abstract

Purpose: Surgical performance depends not only on surgeons' technical skills but also on team communication within and across the different professional groups present during the operation. Therefore, automatically identifying team communication in the OR is crucial for patient safety and advances in the development of computer-assisted surgical workflow analysis and intra-operative support systems. To take the first step, we propose a new task of detecting communication briefings involving all OR team members, i.e. the team Time-out and the StOP?-protocol, by localizing their start and end times in video recordings of surgical operations. Methods: We generate an OR dataset of real surgeries, called Team-OR, with more than one hundred hours of surgical videos captured by the multi-view camera system in the OR. The dataset contains temporal annotations of 33 Time-out and 22 StOP?-protocol activities in total. We then propose a novel group activity detection approach, where we encode both scene context and action features, and use an efficient neural network model to output the results. Results: The experimental results on the Team-OR dataset show that our approach outperforms existing state-of-the-art temporal action detection approaches. It also demonstrates the lack of research on group activities in the OR, proving the significance of our dataset. Conclusion: We investigate the Team Time-Out and the StOP?-protocol in the OR, by presenting the first OR dataset with temporal annotations of group activities protocols, and introducing a novel group activity detection approach that outperforms existing approaches. Code is available at https://github.com/CAMMA-public/Team-OR.

Figures

Figures reproduced from arXiv: 2502.08299 by the authors.

Figure 1
Figure 1. Overview of the Team-OR dataset, consisting of synchronized three ceiling camera views and one laparoscopic view. We blurred the half bodies of the [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The duration distribution of the videos, “Time-out” and “StOP?”. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Examples of the Time-out and StOP?-protocol activities in the dataset. We blurred the half-bodies of the team for privacy concerns. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Framework of our approach. We extract temporal scene context and skeleton features through pretrained VideoMAEv2 [ [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. EgoExOR: An Ego-Exo-Centric Operating Room Dataset for Surgical Activity Understanding

    cs.CV 2025-05 conditional novelty 7.0 of 10

    EgoExOR is a new multimodal, multi-perspective OR dataset with 84,553 annotated frames, plus a benchmark showing that fusing egocentric and exocentric signals improves surgical scene graph generation.

Reference graph

Works this paper leans on

54 extracted references · 38 canonical work pages · cited by 1 Pith paper

  1. [1]

    Layer normalization

    Ba, J.L., 2016. Layer normalization. arXiv preprint arXiv:1607.06450 doi:10.48550/arXiv.1607.06450

    arXiv doi:10.48550/arxiv.1607.06450 2016

  2. [2]

    Parsing human skeletons in an operating room

    Belagiannis, V ., Wang, X., Shitrit, H.B.B., Hashimoto, K., Stauder, R., Aoki, Y ., Kranzfelder, M., Schneider, A., Fua, P., Ilic, S., et al., 2016. Parsing human skeletons in an operating room. Machine Vision and Applications 27, 1035–1046. doi: 10.1007/s00138-016-0792-4

    work page doi:10.1007/s00138-016-0792-4 2016

  3. [3]

    Beldi, G., Bisch-Knaden, S., Banz, V ., M ¨uhlemann, K., Candinas, D.,

    work page

  4. [4]

    Soft-nms– improving object detection with one line of code, in: Proceedings of the IEEE international conference on computer vision, pp

    Bodla, N., Singh, B., Chellappa, R., Davis, L.S., 2017. Soft-nms– improving object detection with one line of code, in: Proceedings of the IEEE international conference on computer vision, pp. 5561–5569. doi:10.1109/ICCV.2017.593

    doi:10.1109/iccv.2017.593 2017

  5. [5]

    ViTALS: Vision Transformer for Action Localization in Surgical Nephrectomy

    Chandra, S., Chowdhury, S.S., Yong, C., Sundaram, C.P., Roy, K., 2024. Vitals: Vision transformer for action localization in surgical nephrec- tomy. arXiv preprint arXiv:2405.02571 doi: 10.48550/arXiv.2405. 02571

    work page Pith review arXiv doi:10.48550/arxiv.2405.02571 2024

  6. [6]

    Dergachyova, O., Bouget, D., Huaulm ´e, A., Morandi, X., Jannin, P.,

    work page

  7. [7]

    Pyskl: Towards good practices for skeleton action recognition, in: Proceedings of the 30th ACM International Conference on Multimedia, pp

    Duan, H., Wang, J., Chen, K., Lin, D., 2022a. Pyskl: Towards good practices for skeleton action recognition, in: Proceedings of the 30th ACM International Conference on Multimedia, pp. 7351–7354. doi:10. 1145/3503161.3548546

    work page arXiv

  8. [8]

    Revisiting skeleton-based action recognition, in: Proceedings of the IEEE /CVF Conference on Computer Vision and Pattern Recognition, pp

    Duan, H., Zhao, Y ., Chen, K., Lin, D., Dai, B., 2022b. Revisiting skeleton-based action recognition, in: Proceedings of the IEEE /CVF Conference on Computer Vision and Pattern Recognition, pp. 2969–

    work page

Show all 54 references
  1. [9]

    Yolox: Exceeding yolo series in 2021

    Ge, Z., Liu, S., Wang, F., Li, Z., Sun, J., 2021. Yolox: Exceeding yolo series in 2021. arXiv preprint arXiv:2107.08430 doi: 10.48550/ arXiv.2107.08430

    2021 doi arXiv

  2. [10]

    Fusing infor- mation from multiple 2d depth cameras for 3d human pose estimation in the operating room

    Hansen, L., Siebert, M., Diesel, J., Heinrich, M.P., 2019. Fusing infor- mation from multiple 2d depth cameras for 3d human pose estimation in the operating room. International journal of computer assisted radiology and surgery 14, 1871–1879. doi:10.1007/s11548-019-02044-7

    2019 doi

  3. [11]

    Act-net: anchor-context action detection in surgery videos, in: International Conference on Medical Image Computing and Computer-Assisted Intervention, Springer

    Hao, L., Hu, Y ., Lin, W., Wang, Q., Li, H., Fu, H., Duan, J., Liu, J., 2023. Act-net: anchor-context action detection in surgery videos, in: International Conference on Medical Image Computing and Computer-Assisted Intervention, Springer. pp. 196–206. doi:10.1007/ 978-3-031-4...

    2023

  4. [12]

    A surgical safety checklist to reduce morbidity and mor- tality in a global population

    Haynes, A.B., Weiser, T.G., Berry, W.R., Lipsitz, S.R., Breizat, A.H.S., Dellinger, E.P., Herbosa, T., Joseph, S., Kibatala, P.L., Lapitan, M.C.M., et al., 2009. A surgical safety checklist to reduce morbidity and mor- tality in a global population. New England journal of medi...

    2009 doi

  5. [13]

    time out

    Johnston, G., Ekert, L., Pally, E., 2009. Surgical site signing and “time out”: issues of compliance or complacence. JBJS 91, 2577–2580. doi:10.2106/JBJS.H.01615

    2009 doi

  6. [14]

    Stop? ii trial: cluster randomized clinical trial to test the implementation of a toolbox for structured communication in the operating room—study protocol

    Keller, S., Tschan, F., Semmer, N.K., Trelle, S., Manser, T., Beldi, G., 2022. Stop? ii trial: cluster randomized clinical trial to test the implementation of a toolbox for structured communication in the operating room—study protocol. Trials 23, 878. doi: 10.1186/ s13063-022-06775-y

    2022

  7. [15]

    Leveraging tempo- ral contextualization for video action recognition, in: European Con- ference on Computer Vision, Springer

    Kim, M., Han, D., Kim, T., Han, B., 2025. Leveraging tempo- ral contextualization for video action recognition, in: European Con- ference on Computer Vision, Springer. pp. 74–91. doi: 10.1007/ 978-3-031-72664-4_5

    2025

  8. [16]

    Impact of team familiarity in the operating room on surgical complications

    Kurmann, A., Keller, S., Tschan-Semmer, F., Seelandt, J., Semmer, N.K., Candinas, D., Beldi, G., 2014. Impact of team familiarity in the operating room on surgical complications. World journal of surgery 38, 3047–3052. doi: 10.1007/s00268-014-2680-2

    2014 doi

  9. [17]

    Learning salient boundary feature for anchor-free temporal action localization, in: Proceedings of the IEEE /CVF Con- ference on Computer Vision and Pattern Recognition, pp

    Lin, C., Xu, C., Luo, D., Wang, Y ., Tai, Y ., Wang, C., Li, J., Huang, F., Fu, Y ., 2021. Learning salient boundary feature for anchor-free temporal action localization, in: Proceedings of the IEEE /CVF Con- ference on Computer Vision and Pattern Recognition, pp. 3320–3329. d...

    2021

  10. [18]

    Bmn: Boundary- matching network for temporal action proposal generation, in: Proceed- ings of the IEEE international conference on computer vision, pp

    Lin, T., Liu, X., Li, X., Ding, E., Wen, S., 2019. Bmn: Boundary- matching network for temporal action proposal generation, in: Proceed- ings of the IEEE international conference on computer vision, pp. 3889–

    2019

  11. [19]

    Surgplan: Surgical phase localization network for phase recognition, in: 2024 IEEE International Symposium on Biomedical Imaging (ISBI), IEEE

    Luo, X., Pang, Y ., Chen, Z., Wu, J., Zhang, Z., Lei, Z., Liu, H., 2024. Surgplan: Surgical phase localization network for phase recognition, in: 2024 IEEE International Symposium on Biomedical Imaging (ISBI), IEEE. pp. 1–5. doi: 10.1109/ISBI56570.2024.10635466

    2024

  12. [20]

    Or black box and surgical control tower: recording and streaming data and analytics to improve surgi- cal care

    Mascagni, P., Padoy, N., 2021. Or black box and surgical control tower: recording and streaming data and analytics to improve surgi- cal care. Journal of Visceral Surgery 158, S18–S25. doi: 10.1016/j. jviscsurg.2021.01.004

    2021 doi

  13. [21]

    Mosaic: a web-based plat- form for collaborative medical video assessment and annotation

    Mazellier, J.P., Boujon, A., Bour-Lang, M., Erharhd, M., Waechter, J., Wernert, E., Mascagni, P., Padoy, N., 2023. Mosaic: a web-based plat- form for collaborative medical video assessment and annotation. arXiv doi:10.48550/arXiv.2312.08593

    2023 doi arXiv

  14. [22]

    Surgical team behaviors and patient outcomes

    Mazzocco, K., Petitti, D.B., Fong, K.T., Bonacum, D., Brookey, J., Gra- ham, S., Lasky, R.E., Sexton, J.B., Thomas, E.J., 2009. Surgical team behaviors and patient outcomes. The American journal of surgery 197, 678–685. doi: 10.1016/j.amjsurg.2008.03.002

    2009 doi

  15. [23]

    Overdyk, F.J., Dowling, O., Newman, S., Glatt, D., Chester, M., Armellino, D., Cole, B., Landis, G.S., Schoenfeld, D., DiCapua, J.F.,

  16. [24]

    ¨Ozsoy, E., Czempiel, T., Holm, F., Pellegrini, C., Navab, N., 2023. Labrad-or: lightweight memory scene graphs for accurate bimodal reasoning in dynamic operating rooms, in: International Conference on Medical Image Computing and Computer-Assisted Intervention, Springer. pp. ...

    2023 doi

  17. [25]

    ¨Ozsoy, E., Czempiel, T., ¨Ornek, E.P., Eck, U., Tombari, F., Navab, N.,

  18. [26]

    4d-or: Semantic scene graphs for or domain model- ing, in: International Conference on Medical Image Computing and Computer-Assisted Intervention, Springer

    ¨Ozsoy, E., ¨Ornek, E.P., Eck, U., Czempiel, T., Tombari, F., Navab, N., 2022. 4d-or: Semantic scene graphs for or domain model- ing, in: International Conference on Medical Image Computing and Computer-Assisted Intervention, Springer. pp. 475–485. doi:10.1007/ 978-3-031-16449-1_45

    2022

  19. [27]

    Machine and deep learning for workflow recognition during surgery

    Padoy, N., 2019. Machine and deep learning for workflow recognition during surgery. Minimally Invasive Therapy & Allied Technologies 28, 82–90. doi: 10.1080/13645706.2019.1584116

    2019

  20. [28]

    BMJ quality & safety 25, 947–953

    Remote video auditing with real-time feedback in an academic surgical suite improves safety and e fficiency metrics: a cluster ran- domised study. BMJ quality & safety 25, 947–953. doi: 10.1136/ bmjqs-2015-004226

    2015

  21. [29]

    Sˆ 2former-or: Single-stage bimodal transformer for scene graph gener- ation in or

    Pei, J., Guo, D., Zhang, J., Lin, M., Jin, Y ., Heng, P.A., 2024. Sˆ 2former-or: Single-stage bimodal transformer for scene graph gener- ation in or. arXiv preprint arXiv:2402.14461 doi: 10.48550/arXiv. 2402.14461

    2024 doi arXiv

  22. [30]

    Real- time segmentation and recognition of surgical tasks in cataract surgery videos

    Quellec, G., Lamard, M., Cochener, B., Cazuguel, G., 2014. Real- time segmentation and recognition of surgical tasks in cataract surgery videos. IEEE transactions on medical imaging 33, 2352–2360. doi: 10. 1109/TMI.2014.2340473

    2014

  23. [31]

    Focal loss for dense object detection, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- 8 tern Recognition, pp

    Ross, T.Y ., Doll´ar, G., 2017. Focal loss for dense object detection, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- 8 tern Recognition, pp. 2980–2988. doi: 10.1109/ICCV.2017.324

    2017 doi

  24. [32]

    Assessing distractors and teamwork during surgery: developing an event-based method for di- rect observation

    Seelandt, J.C., Tschan, F., Keller, S., Beldi, G., Jenni, N., Kurmann, A., Candinas, D., Semmer, N.K., 2014. Assessing distractors and teamwork during surgery: developing an event-based method for di- rect observation. BMJ quality & safety 23, 918–929. doi: 10.1136/ bmjqs-2014-002860

    2014

  25. [34]

    The who safer surgery checklist time out procedure revisited: Strategies to optimise compliance and safety

    Papadakis, M., Meiwandi, A., Grzybowski, A., 2019. The who safer surgery checklist time out procedure revisited: Strategies to optimise compliance and safety. International Journal of Surgery 69, 19–22. doi:10.1016/j.ijsu.2019.07.006

    2019 doi

  26. [35]

    Unsupervised domain adap- tation for clinician pose estimation and instance segmentation in the op- erating room

    Srivastav, V ., Gangi, A., Padoy, N., 2022. Unsupervised domain adap- tation for clinician pose estimation and instance segmentation in the op- erating room. Medical image analysis 80, 102525. doi: 10.1016/j. media.2022.102525

    2022

  27. [36]

    Mvor: A multi-view rgb-d operating room dataset for 2d and 3d human pose estimation

    Srivastav, V ., Issenhuth, T., Abdolrahim, K., de Mathelin, M., Gangi, A., Padoy, N., 2018. Mvor: A multi-view rgb-d operating room dataset for 2d and 3d human pose estimation. doi: 10.48550/arXiv.1808. 08180

    2018 doi

  28. [37]

    Deep high-resolution rep- resentation learning for human pose estimation, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Sun, K., Xiao, B., Liu, D., Wang, J., 2019. Deep high-resolution rep- resentation learning for human pose estimation, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 5693–5703. doi: 10.1109/CVPR.2019.00584

    2019

  29. [38]

    Temporalmaxer: Maximize tem- poral context with only max pooling for temporal action localization

    Tang, T.N., Kim, K., Sohn, K., 2023. Temporalmaxer: Maximize tem- poral context with only max pooling for temporal action localization. arXiv preprint arXiv:2303.09055 doi:10.48550/arXiv.2303.09055

    2023 doi arXiv

  30. [39]

    Tschan, F., Keller, S., Semmer, N.K., Timm-Holzer, E., Zimmermann, J., Huber, S.A., Wrann, S., H ¨ubner, M., Banz, V ., Prevost, G.A., et al.,

  31. [40]

    Two-stream adaptive graph convolutional networks for skeleton-based action recognition, in: Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Shi, L., Zhang, Y ., Cheng, J., Lu, H., 2019. Two-stream adaptive graph convolutional networks for skeleton-based action recognition, in: Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 12026–12035. doi: 10.1109/CVPR.2019.01230

    2019

  32. [41]

    V olkov, M., Hashimoto, D.A., Rosman, G., Meireles, O.R., Rus, D.,

  33. [42]

    Videomae v2: Scaling video masked autoen- coders with dual masking, in: Proceedings of the IEEE /CVF Confer- ence on Computer Vision and Pattern Recognition, pp

    Wang, L., Huang, B., Zhao, Z., Tong, Z., He, Y ., Wang, Y ., Wang, Y ., Qiao, Y ., 2023. Videomae v2: Scaling video masked autoen- coders with dual masking, in: Proceedings of the IEEE /CVF Confer- ence on Computer Vision and Pattern Recognition, pp. 14549–14560. doi:10.1109/C...

    2023

  34. [43]

    G-tad: Sub-graph localization for temporal action detection, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogni- tion, pp

    Xu, M., Zhao, C., Rojas, D.S., Thabet, A., Ghanem, B., 2020. G-tad: Sub-graph localization for temporal action detection, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogni- tion, pp. 10156–10165. doi: 10.1109/CVPR42600.2020.01017

    2020

  35. [44]

    Actionformer: Localizing moments of actions with transformers, in: European Conference on Computer Vision, Springer

    Zhang, C.L., Wu, J., Li, Y ., 2022a. Actionformer: Localizing moments of actions with transformers, in: European Conference on Computer Vision, Springer. pp. 492–510. doi: 10.1007/978-3-031-19772-7_ 29

    doi

  36. [45]

    Bytetrack: Multi-object tracking by associating every detection box, in: European conference on computer vision, Springer

    Zhang, Y ., Sun, P., Jiang, Y ., Yu, D., Weng, F., Yuan, Z., Luo, P., Liu, W., Wang, X., 2022b. Bytetrack: Multi-object tracking by associating every detection box, in: European conference on computer vision, Springer. pp. 1–21. doi: 10.1007/978-3-031-20047-2_1

    doi

  37. [46]

    Distance-iou loss: Faster and better learning for bounding box regression, in: Pro- ceedings of the AAAI conference on artificial intelligence, pp

    Zheng, Z., Wang, P., Liu, W., Li, J., Ye, R., Ren, D., 2020. Distance-iou loss: Faster and better learning for bounding box regression, in: Pro- ceedings of the AAAI conference on artificial intelligence, pp. 12993– 13000. doi: 10.1609/aaai.v34i07.6999

    2020 doi

  38. [48]

    Van Klei, W., Ho ff, R., Van Aarnhem, E., Simmermacher, R., Regli, L., Kappen, T., Van Wolfswinkel, L., Kalkman, C., Buhre, W., Peelen, L.,

  39. [144]

    doi: 10.1093/bjs/znab384

    doi

  40. [2012]

    surgical safety checklist

    E ffects of the introduction of the who “surgical safety checklist” on in-hospital mortality: a cohort study. Annals of surgery 255, 44–49. doi:10.1097/SLA.0b013e31823779ae

    doi

  41. [2016]

    International journal of computer assisted radiology and surgery 11, 1081–1089

    Automatic data-driven real-time segmentation and recognition of surgical workflow. International journal of computer assisted radiology and surgery 11, 1081–1089. doi:10.1007/s11548-016-1371-x

    doi

  42. [2017]

    Machine learning and coresets for automated real-time video segmentation of laparoscopic and robot-assisted surgery, in: 2017 IEEE international conference on robotics and automation (ICRA), IEEE. pp. 754–759. doi: 10.1109/ICRA.2017.7989093

    2017

  43. [2022]

    British Journal of Surgery 109, 136–

    Effects of structured intraoperative briefings on patient outcomes: multicentre before-and-after study. British Journal of Surgery 109, 136–

  44. [2024]

    International Journal of Computer Assisted Radiology and Surgery 19, 791–799

    Holistic or domain modeling: a semantic scene graph approach. International Journal of Computer Assisted Radiology and Surgery 19, 791–799. doi: 10.1007/s11548-023-03022-w

    doi

  45. [2978]

    doi: 10.1109/CVPR52688.2022.00298

    2022

  46. [3898]

    doi: 10.1109/ICCV.2019.00399

    2019

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.