List of Papers Browse by Subject Areas Author List
Abstract
The advent of telemedicine represents a transformative development in leveraging technology to extend the reach of specialized medical expertise to remote surgeries, a field where the immediacy of expert guidance is paramount. However, the intricate dynamics of Operating Room (OR) scene pose unique challenges for telemedicine, particularly in achieving high-fidelity, real-time scene reconstruction and transmission amidst obstructions and bandwidth limitations. This paper introduces TeleOR, a pioneering system designed to address these challenges through real-time OR scene reconstruction for Tele-intervention. TeleOR distinguishes itself with three innovative approaches: dynamic self-calibration, which leverages inherent scene features for calibration without the need for preset markers, allowing for obstacle avoidance and real-time camera adjustment; selective OR reconstruction, focusing on dynamically changing scene segments to reduce reconstruction complexity; and viewport-adaptive transmission, optimizing data transmission based on real-time client feedback to efficiently deliver high-quality 3D reconstructions within bandwidth constraints. Comprehensive experiments on the 4D-OR surgical scene dataset demostrate the superiority and applicability of TeleOR, illuminating the potential to revolutionize tele-interventions by overcoming the spatial and technical barriers inherent in remote surgical guidance.
Links to Paper and Supplementary Materials
Main Paper (Open Access Version): https://papers.miccai.org/miccai-2024/paper/0757_paper.pdf
SharedIt Link: pending
SpringerLink (DOI): pending
Supplementary Material: N/A
Link to the Code Repository
N/A
Link to the Dataset(s)
N/A
BibTex
@InProceedings{Wu_TeleOR_MICCAI2024,
author = { Wu, Yixuan and Hu, Kaiyuan and Shao, Qian and Chen, Jintai and Chen, Danny Z. and Wu, Jian},
title = { { TeleOR: Real-time Telemedicine System for Full-Scene Operating Room } },
booktitle = {proceedings of Medical Image Computing and Computer Assisted Intervention -- MICCAI 2024},
year = {2024},
publisher = {Springer Nature Switzerland},
volume = {LNCS 15006},
month = {October},
page = {pending}
}
Reviews
Review #1
- Please describe the contribution of the paper
The authors propose a system for OR scene reconstruction and real-time streaming of the reconstructed scene over a network. The system has features for adaptive bandwidth and also viewport adaptive transmission features.
- Please list the main strengths of the paper; you should write about a novel formulation, an original way to use data, demonstration of clinical feasibility, a novel application, a particularly strong evaluation, or anything else that is a strong aspect of this work. Please provide details, for instance, if a method is novel, explain what aspect is novel and why this is interesting.
The main strengths of the paper are that the authors are using an open dataset for validation, the method does not need frame markers present in the scene for calibration purposes. Also the method runs on a Jetson Nano device thus it is suitable for a lightweight device that can be easily placed in the OR.
- Please list the main weaknesses of the paper. Please provide details, for instance, if you think a method is not novel, explain why and provide a reference to prior work.
My main criticism is that the paper is not validated by using clinical personnel that are supposed to use this system. Further the authors do not provide validation against known methods or discuss their results vs. the ones published in the literature. The novelty of the technical implementations are rather limited and should be compared against state of the art methods (NeRF or Gaussian splats).
- Please rate the clarity and organization of this paper
Good
- Please comment on the reproducibility of the paper. Please be aware that providing code and data is a plus, but not a requirement for acceptance.
The submission does not provide sufficient information for reproducibility.
- Do you have any additional comments regarding the paper’s reproducibility?
Camera setup and implementation details are missing. The method is presented at a very high level no detail are given.
- Please provide detailed and constructive comments for the authors. Please also refer to our Reviewer’s guide on what makes a good review. Pay specific attention to the different assessment criteria for the different paper categories (MIC, CAI, Clinical Translation of Methodology, Health Equity): https://conferences.miccai.org/2024/en/REVIEWER-GUIDELINES.html
The method can be useful in a modern OR scenario, however the 3D reconstructions should be presented to clinical personnel for a qualitative validation. The reconstruction of the overall scene is presented however more focus should be given to specific elements in the scene (e.g. surgical area).
- Rate the paper on a scale of 1-6, 6 being the strongest (6-4: accept; 3-1: reject). Please use the entire range of the distribution. Spreading the score helps create a distribution for decision-making
Reject — should be rejected, independent of rebuttal (2)
- Please justify your recommendation. What were the major factors that led you to your overall score for this paper?
-limited technical novelty -no validation against state of the art -no validation of clinical personnel
- Reviewer confidence
Very confident (4)
- [Post rebuttal] After reading the author’s rebuttal, state your overall opinion of the paper if it has been changed
N/A
- [Post rebuttal] Please justify your decision
N/A
Review #2
- Please describe the contribution of the paper
The paper describes a method to enable reconstructions of an operating room with multiple RGB-D sensors and a compression scheme to stream the resulting reconstruction over bandwidth-limited network connections. The RGB-D cameras can move, and the system self-calibrates their camera parameters based on natural features in the operating room. Furthermore, the system detects areas in each RGB-D stream that stay mostly constant over time through optical flow to only transmit parts of the streams that change meaningfully. Another aspect of the compression scheme is that only parts currently being viewed by a remote viewer are being transmitted, predicting a user’s viewpoint in a VR HMD and transmitting the center of the FoV with higher quality than the periphery. The implementation is tested on the publicly available 4D-OR dataset, and the authors report visual quality metrics of the resulting images and rendering performance with multiple limited bandwidths.
- Please list the main strengths of the paper; you should write about a novel formulation, an original way to use data, demonstration of clinical feasibility, a novel application, a particularly strong evaluation, or anything else that is a strong aspect of this work. Please provide details, for instance, if a method is novel, explain what aspect is novel and why this is interesting.
the paper is well written, without major spelling or grammar flaws. The literature review is extensive and discusses a large amount of related papers. The reconstruction, self-calibration and compression scheme seems well thought out and seems to work well for limited bandwidths. The bandwidth requirements make the system very attractive for real world use cases, as for hospitals in rural areas (that would benefit the most from telepresence in the OR) may be on very bandwidth limited network connections. The objective metrices for the evaluation seem appropriate.
- Please list the main weaknesses of the paper. Please provide details, for instance, if you think a method is not novel, explain why and provide a reference to prior work.
My main complaint of the paper is, that I am confused of the nature of the contribution. Is TeleOR a physical system comprising of cameras in a room, mounted on the ceiling and some users somewhere else how can join the reconstruction in real time with head mounted displays. Or is it an algorithm or software framework, without a physical aspect to it. Please provide a clear explanation of whether TeleOR involves physical hardware, operates entirely in software, or utilizes a virtual simulation environment. The authors consistently mention experiments on the 4D-OR dataset, which is publicly available. All figures show the scene of the 4D-OR dataset. However, the cameras in the 4D-OR dataset do not move, whereas the authors of this paper mention moving platforms for the cameras.Or are the experiments done in a virtual environment, with cameras on mobiile platforms in a simulation based on the 4D or dataset? Either variant is completely fine, and with the promising results can be published. But the authors should state the nature of their system clearly.
Another aspect which I can be improved is the literature review. The number of papers is fine, however in the text I found the association between claims and corresponding citations confusing. Moreover some references I think are not closely related, whereas a few may be missing (see detailed comments below).
- Please rate the clarity and organization of this paper
Good
- Please comment on the reproducibility of the paper. Please be aware that providing code and data is a plus, but not a requirement for acceptance.
The submission does not mention open access to source code or data but provides a clear and detailed description of the algorithm to ensure reproducibility.
- Do you have any additional comments regarding the paper’s reproducibility?
The system presented here seems quite complex and does not comprise a single algorithm, etc, which can be implemented in a few days. Therefore, I deeply hope that the authors are considering making their implementation available to other researchers. No mention of open-sourcing is made in the text. I think all the necessary details are in the text to replicate the system, but building a similar system would be a huge engineering task.
- Please provide detailed and constructive comments for the authors. Please also refer to our Reviewer’s guide on what makes a good review. Pay specific attention to the different assessment criteria for the different paper categories (MIC, CAI, Clinical Translation of Methodology, Health Equity): https://conferences.miccai.org/2024/en/REVIEWER-GUIDELINES.html
I will give detailed feedback to each section of the paper below:
Title I like the name “TeleOR”. But the subtitle is quite empty, and from it alone, the reader can not understand what the paper is about. It’s not wrong, just not descriptive. Therefore, I suggest finding a more descriptive title.
Abstract “the real-world 4D-OR surgical scene dataset” - yes, this dataset was captured with real cameras, but it does not show real surgeries. They were ‘simulated’ by actors. Therefore, the term ‘real-world’ may be misleading.
Keywords I don’t know what “Surgical understanding” is. I suggest a different keyword, maybe in the direction of bandwidth adaptive 3D reconstruction compression, which I find a very interesting aspect of the paper.
Introduction -“The deployment of telemedicine systems necessitates a series of critical steps, including onsite scene reconstruction, transmission, remote rendering, and intervention.” - The authors might want to define what exactly they define as telemedicine. While the version of telemedicine, that the authors discuss includes all of these steps, some readers might consider a videoconference call with a patient or colleague as telemedicine, which does not require these technical steps. -“Previous research has explored various methods employing multiple static cameras to capture.” - the paper [1] that is referenced here talks about RGBD cameras, and the 4D-OR dataset uses RGBD as well. Therefore, the authors might want to clarify that these cameras are special 3d reconstruction cameras, even though 3d reconstruction is possible with multiple cameras, e.g., multiple view reconstruction. -“neural implicit functions [8,13,23,25]” - I’m not a fan of listing many references just to back up one claim without discussing the differences of the different cited works. Consider describing these in more detail or focusing on the important works. Overall, the authors may consider moving the references around or being more explicit (e.g., Author1 et al. [1] did X, while Author2 et al. [2] did Y, etc.) to make sure the reader understands which statement is backed up by which reference. Also, if the authors cite other papers about neural rendering, one that is currently popular and the authors might want to consider adding is “Mildenhall, B., Srinivasan, P. P., Tancik, M., Barron, J. T., Ramamoorthi, R., & Ng, R. (2021). Nerf: Representing scenes as neural radiance fields for view synthesis.” -“However, this simplification tends” - should be in plural. -“may restrict the number of remote clients who can access the system” - is this a problem brought up by some authors, ideally clinical experts? I have difficulty coming up with a good example where teleconsultation from multiple remote users would be clinically necessary. -“The Operating Room (OR) presents a dynamic and complex scenario, teeming with medical staff, patients, and an array of medical equipment [18].” - While it is nice to cite the authors of the used dataset also with their other papers, THIS particular statement does not really motivate including a reference. This is the only instance of citing [18]. Either it makes sense to back up another claim with this reference or remove it. -What are “traditional static settings”? -“To address these challenges, in this work, we propose TeleOR, the first real-time Operating Room scene reconstruction system for Tele-intervention.” - I disagree. This is not the first system to reconstruct an operating room and have remote users join. One earlier (probably still not the first) counter-example I know would be K. Yu et al., “Magnoramas: Magnifying Dioramas for Precise Annotations in Asymmetric 3D Teleconsultation,” 2021 IEEE Virtual Reality and 3D User Interfaces. Coincidentally, this publication seems to have been conducted in the same environment as the 4D-OR dataset. -some definite articles (the) and indefinite articles (a, an) appear to be missing in the text.
Related Work -“Nevertheless, current approaches have yet to achieve real-time reconstruction” - it is unclear, what exactly the authors mean by “reconstruction”. Moreover, the above-mentioned paper by Yu et al. displays multiple RGBD streams in an immersive VR telepresence to a remote user in real-time. -I don’t really understand how the paragraph about “Surgical Intervention” lists any related work. The title of the paragraph is misleading, and I fail to understand how the works are related to the proposed system. e.g., reference [3] reconstructs an abdomen from laparoscopic views, and I don’t see how this is related to room reconstruction with RGB-D cameras. Either the authors should introduce this topic in more detail so that a reader understands how it relates to this system, or remove this paragraph.
Methodology -The sentence “In Fig. 1, our proposed TeleOR systematically integrates” is weird. Fig. 1 shows how TeleOR does it.
- “medical stuff” -> ‘staff’ -“impacting the calibration and synchronization” - how is synchronization affected? The Kinects used in 4D-OR were synchronized using sync cables. -“Restriction of pre-recording setups” seems like an unfitting umbrella term for the explanation that follows. -Footnote 1 could be moved to the regular text. -“due to shifts in the camera’s FoV” - the FoV is fixed for the camera? Do the authors mean repositioning the camera? -“camera’s calibration parameters, ensuring accurate and synchronized” - how does camera calibration affect synchronization? To me, these two concepts seem unrelated. -“these partially masked frames are forwarded to the synthesis server” - I think this is the first mention of a server. So far, no word on software architecture has been mentioned. The authors may consider stating their software architecture further in the paper. -“on the user’s past Field of View (FoV) movements to construct “ - the abbreviation was introduced before.
Experiment -“The supplementary materials provide a detailed description of the rendering data format” - I know that the authors are running out of space. However, I think that the data format would be better suited as a regular figure in the paper than as suppl.mat. -I don’t think footnote 2 is important. I think readers of a telepresence system paper can be expected to know about their bandwidth requirements. -“with FPS reflecting the smoothness of video playback” - what do the authors understand as a “frame”? is it the final 2D image displayed to the user, or is it one “frame” of the 3d reconstruction? -footnote 3 is unnecessary. I think readers of a telepresence system paper can judge what kind of framerate they would want to achieve, be it 24 like in a movie, or rather 60 or 90 in an HMD, etc.
References -For reference 10 the authors include the DOI twice -Only two references include the DOI. The authors may consider listing the DOI of all references that have one assigned.
- Reference 20 is the same paper as 19.
- Rate the paper on a scale of 1-6, 6 being the strongest (6-4: accept; 3-1: reject). Please use the entire range of the distribution. Spreading the score helps create a distribution for decision-making
Weak Accept — could be accepted, dependent on rebuttal (4)
- Please justify your recommendation. What were the major factors that led you to your overall score for this paper?
I think the presented technique would help many implementers of telepresence systems that want to transmit multiple RGB-D streams. Therefore, I would like to see this idea published, ideally with the source code attached. However, the paper has a few flaws that should be addressed before it is published. I was especially confused about the fact that they only use the 4D-OR dataset, but then, on the other hand, they seem to have another separate 3D reconstruction environment with cameras, servers, and HMDs.
- Reviewer confidence
Confident but not absolutely certain (3)
- [Post rebuttal] After reading the author’s rebuttal, state your overall opinion of the paper if it has been changed
Weak Accept — could be accepted, dependent on rebuttal (4)
- [Post rebuttal] Please justify your decision
While I agree with the other reviewers that the paper has some weaknesses, I think the authors addressed the reviewer’s concerns well in their rebuttal and I agree with the author’s statements. They also show a willingness to make suggested modifications to their paper. I hope the authors follow through with their plan to make their implementation available, as I would be interested in exact comparisons to other systems down the line.
In conclusion, I think the paper is just above the bar for acceptance. Therefore, I maintain my recommendation at ‘weak accept’.
Rank 2 of 3
Review #3
- Please describe the contribution of the paper
The paper introduces several techniques that taken together enable for alleged high quality near real-time streaming of operating room geometry. The techniques are a dynamic self-calibration muticamera method, selective reconstruction and a viewport-adaptive transmission strategy.
- Please list the main strengths of the paper; you should write about a novel formulation, an original way to use data, demonstration of clinical feasibility, a novel application, a particularly strong evaluation, or anything else that is a strong aspect of this work. Please provide details, for instance, if a method is novel, explain what aspect is novel and why this is interesting.
-This paper presents an elegant and innovative solution to a difficult problem. -The methods are simple yet clever. -The novelty is high. -The potential clinical benefits could be high.
- Please list the main weaknesses of the paper. Please provide details, for instance, if you think a method is not novel, explain why and provide a reference to prior work.
-The text is too verbose and redundant in some places. -The method is claimed to target (or be of potential use for) underdeveloped regions, but then later on, authors assess that hospitals typically have available bandwidth in 100 Mbps to 1Gbps range. These two statements are hard to reconcile. Hospitals are also typically going to have a lot of data flowing in and out in order to use all other systems deployed in it, so using a significant portion of the total available bandwidth for one tele-operated procedure might not be realistic. The “limited bandwidth” scenario presented in the paper is therefore likely much less stringent than the authors claim. Consequently it would have been interesting to present a fourth scenario with even lower bandwidth, perhaps 5 Mbps or even 1. -Too much focus is placed on one of the two evaluation metrics relative to the other. While I agree with the authors that a framerate of 22 fps is good, a latency of 300 milliseconds is problematic. It can also be hypothesized that if the “very limited” scenario suggested above had been implemented, then the latency would have been quite a bit higher still. This kind of latency could hinder system use to some extent. -Reconstruction quality is only assessed in 2D renderings. It would have been interesting to also provide a 3D measure.
- Please rate the clarity and organization of this paper
Very Good
- Please comment on the reproducibility of the paper. Please be aware that providing code and data is a plus, but not a requirement for acceptance.
The submission does not provide sufficient information for reproducibility.
- Do you have any additional comments regarding the paper’s reproducibility?
The method is not easily reproducible.
- Please provide detailed and constructive comments for the authors. Please also refer to our Reviewer’s guide on what makes a good review. Pay specific attention to the different assessment criteria for the different paper categories (MIC, CAI, Clinical Translation of Methodology, Health Equity): https://conferences.miccai.org/2024/en/REVIEWER-GUIDELINES.html
The collection of methods presented are very relevant to the field and timely. While this work is still at an early stage in development, it looks very promising. I strongly encourage authors to expand on this early version and submit to a journal, regardless of MICCAI’s acceptance decision.
- Rate the paper on a scale of 1-6, 6 being the strongest (6-4: accept; 3-1: reject). Please use the entire range of the distribution. Spreading the score helps create a distribution for decision-making
Accept — should be accepted, independent of rebuttal (5)
- Please justify your recommendation. What were the major factors that led you to your overall score for this paper?
The novelty and relevance to this community are both high.
- Reviewer confidence
Very confident (4)
- [Post rebuttal] After reading the author’s rebuttal, state your overall opinion of the paper if it has been changed
Weak Accept — could be accepted, dependent on rebuttal (4)
- [Post rebuttal] Please justify your decision
I much appreciate the author’s effort to answer and the added data to try to justify viability under a more constrained bandwidth scenario. My initial criticism remains however. 300 ms is a large latency, one that would impact performance. The low framerate of the low bandwidth scenario would as well. The negative impact of both of these was shown before (for instance see: Sielhorst, Tobias, et al. “Depth perception–a major issue in medical ar: evaluation study by twenty surgeons.” Medical Image Computing and Computer-Assisted Intervention–MICCAI 2006: 9th International Conference, Copenhagen, Denmark, October 1-6, 2006. Proceedings, Part I 9. Springer Berlin Heidelberg, 2006.) The paper is interesting and would be worth reading, but the performance is overstated.
Author Feedback
We appreciate the reviewers’ valuable comments and are encouraged by their unanimous recognition of TeleOR’s significance in clinical settings. We’ll first address the common issues, and then respond to their specific concerns.
General concern: Implementation details of TeleOR
Answer: Upon acceptance, we’ll open-source the code and make the complete implementations available for supporting the development of this promising topic.
Responses to R1:
Concern 1: Lack of validation by clinical personnel
A1: Due to ethical constraints, we use the public 4D-OR dataset, which simulates OR scenes without real patients and lesions. Thus, remote experts could not assess our TeleOR for actual surgeries in this context. Recognizing the importance of clinical validation, we are now conducting trials with our partner hospital to ensure that our TeleOR meets the clinical standards.
Concern 2: Lack of comparison with SOTA
A2: We don’t compare with common reconstruction methods like NeRF and 3D Gaussian Splatting for two reasons: (1) These methods require substantial computations, resulting in high latency inappropriate for real-time applications, as detailed in our related work discussion; (2) they struggle with alignment from multiple views and require static frame sequences, unsuitable for dynamic OR scenes.
Concern 3: Lack of novelty
A3: We would like to highlight TeleOR’s innovations designed to achieve real-time applications: Selective Reconstruction to reduce computations and Viewport-Adaptive Transmission for optimized data transmission. Moreover, the novelty of TeleOR has been acknowledged by all other reviewers, emphasizing its contributions in practice.
Concern 4: Lack of implementation details (e.g., camera setup details)
A4: We would like to clarify that the implementation details, including the camera setup, are provided in Sec. 4.1. Specifically, the camera array consists of six Microsoft Azure Kinect RGB-D sensors. Additionally, Sec. 3.1 outlines the details of camera initialization and calibration.
Responses to R3
Concern 1: Performance under lower bandwidth
A1: At 5 Mbps: avg. frame rate: 12.7fps, scene reuse ratio: 92.3%, and avg. latency: 321ms. At 1 Mbps: avg. frame rate: 3.6 fps, scene reuse ratio: 96.2%, and avg. latency: 332 ms.
Concern 2: Latency increase with lower bandwidth?
A2: As listed in A1 above, under 5 or 1 Mbps, latency does not increase. This stability is due to:
(1) TeleOR adjusts the scene reuse ratio and point cloud density based on bandwidth to ensure real-time streaming. Thus, limited network conditions mainly affect the frame rate instead of latency.
(2) Latency does not affect visual smoothness but only the time gap perceived by remote doctors. A ~300ms delay is completely manageable within a surgical context.
Concern 3: 3D metrics for reconstruction quality assessment
A3: Due to lack of well-defined 3D quality metrics, in this work, we develop MSSIM for 3D evaluation by assessing the perceived quality from multiple predefined viewpoints. Moreover, as TeleOR presents a 2D image to the user, whether via VR headset or screen, assessing the quality of 2D renderings provides a direct measure.
Responses to R4
Concern 1: The nature of TeleOR’s contributions, i.e., is TeleOR a physical system or an algorithm?
A1: First, TeleOR is a physical system that incorporates hardware components, including depth sensor array, graphic processing unit, and VR headset. It is designed to capture, stream, and display a reconstructed OR scene in real-time. Second, this paper focuses on technical innovations that facilitate the real-time operation of this physical system.
Concern 2: Improvements needed in literature review and detailed suggestions (title, keywords, terminology, etc)
A2: Thank you for your valuable feedback! We’ll refine the literature review for clarity, and incorporate your suggestions throughout this paper into the revised version to ensure that it meets the highest standard.
Meta-Review
Meta-review #1
- After you have reviewed the rebuttal and updated reviews, please provide your recommendation based on all reviews and the authors’ rebuttal.
Accept
- Please justify your recommendation. You may optionally write justifications for ‘accepts’, but are expected to write a justification for ‘rejects’
While R1 expressed concerns about limited technical novelty and validation against existing methods, they acknowledged the relevance of the proposed methods and leaned towards a weak accept post-rebuttal. R3 appreciated the innovation and potential clinical benefits of TeleOR but highlighted issues such as latency issues and mention papers that have showed the impact of these on surgical performance/usability of systems. Similarly, R4 appreciated the well-written paper and novelty of the proposed techniques but expressed confusion about the nature of the contribution and suggested improvements in clarity and reproducibility. Overall, all reviewers recommended acceptance with some reservations, and I believe that with the mentioned improvements in the rebuttal the paper could be acceptable.
- What is the rank of this paper among all your rebuttal papers? Use a number between 1/n (best paper in your stack) and n/n (worst paper in your stack of n papers). If this paper is among the bottom 30% of your stack, feel free to use NR (not ranked).
While R1 expressed concerns about limited technical novelty and validation against existing methods, they acknowledged the relevance of the proposed methods and leaned towards a weak accept post-rebuttal. R3 appreciated the innovation and potential clinical benefits of TeleOR but highlighted issues such as latency issues and mention papers that have showed the impact of these on surgical performance/usability of systems. Similarly, R4 appreciated the well-written paper and novelty of the proposed techniques but expressed confusion about the nature of the contribution and suggested improvements in clarity and reproducibility. Overall, all reviewers recommended acceptance with some reservations, and I believe that with the mentioned improvements in the rebuttal the paper could be acceptable.
Meta-review #2
- After you have reviewed the rebuttal and updated reviews, please provide your recommendation based on all reviews and the authors’ rebuttal.
Accept
- Please justify your recommendation. You may optionally write justifications for ‘accepts’, but are expected to write a justification for ‘rejects’
TeleOR is an advanced system designed for enhancing tele-intervention in OR through real-time, high-fidelity scene reconstructions. By integrating dynamic self-calibration, selective reconstruction, and viewport adaptive transmission, it addresses issues like camera occlusions, scene complexity, and bandwidth constraints, enhancing remote surgical guidance. The evaluation stage of the paper is conducted using the public operating room dataset 4D-OR. . The authors have adequately addressed the major concerns and questions raised by the reviewers regarding the significance of the proposed methods and some of the technical aspects. The main limitation of the paper is a lack of validation across existing methods. I would agree with several reviewers and lean towards a weak accept.
- What is the rank of this paper among all your rebuttal papers? Use a number between 1/n (best paper in your stack) and n/n (worst paper in your stack of n papers). If this paper is among the bottom 30% of your stack, feel free to use NR (not ranked).
TeleOR is an advanced system designed for enhancing tele-intervention in OR through real-time, high-fidelity scene reconstructions. By integrating dynamic self-calibration, selective reconstruction, and viewport adaptive transmission, it addresses issues like camera occlusions, scene complexity, and bandwidth constraints, enhancing remote surgical guidance. The evaluation stage of the paper is conducted using the public operating room dataset 4D-OR. . The authors have adequately addressed the major concerns and questions raised by the reviewers regarding the significance of the proposed methods and some of the technical aspects. The main limitation of the paper is a lack of validation across existing methods. I would agree with several reviewers and lean towards a weak accept.