Paper Info Reviews Author Feedback Meta-Review Back to top
List of Papers Browse by Subject Areas Author List

Abstract

Catheter ablation is a prevalent procedure for treating atrial fibrillation, primarily utilizing catheters equipped with electrodes to gather electrophysiological signals. However, the localization of catheters in fluoroscopy images presents a challenge for clinicians due to the complexity of the intervention processes. In this paper, we propose SIX-Net, a novel algorithm intending to localize landmarks of electrodes in fluoroscopy images precisely, by mixing up spatial-context information from three aspects: First, we propose a new network architecture specially designed for global-local spatial feature aggregation; Then, we mix up spatial correlations between segmentation and landmark detection, by sequential connections between the two tasks with the help of the Segment Anything Model; Finally, a weighted loss function is carefully designed considering the relative spatial-arrangement information among electrodes in the same image. Experiment results on the test set and two clinical-challenging subsets reveal that our method outperforms several state-of-the-art landmark detection methods (~50% improvement for RF and ~25% improvement for CS).

Links to Paper and Supplementary Materials

Main Paper (Open Access Version): https://papers.miccai.org/miccai-2024/paper/2762_paper.pdf

SharedIt Link: https://rdcu.be/dVZez

SpringerLink (DOI): https://doi.org/10.1007/978-3-031-72378-0_32

Supplementary Material: https://papers.miccai.org/miccai-2024/supp/2762_supp.pdf

Link to the Code Repository

N/A

Link to the Dataset(s)

N/A

BibTex

@InProceedings{Wan_SIXNet_MICCAI2024,
        author = { Wang, Xinyi and Xu, Zikang and Zhu, Heqin and Yao, Qingsong and Sun, Yiyong and Zhou, S. Kevin},
        title = { { SIX-Net: Spatial-context Information miX-up for Electrode Landmark Detection } },
        booktitle = {proceedings of Medical Image Computing and Computer Assisted Intervention -- MICCAI 2024},
        year = {2024},
        publisher = {Springer Nature Switzerland},
        volume = {LNCS 15001},
        month = {October},
        page = {338 -- 348}
}

Reviews

Review #1

  • Please describe the contribution of the paper

    Ablation is widely used for clinical treatment of many cardiac diseases including atrial fibrillation. Indeed, detection catheters in fluoroscopy image are a clinically important task. In this study, the authors developed a computational pipeline to detect catheter locations and compare its performance with other existing approaches using three open dataset. It shows the superiority of their approach.

  • 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 has many strengths:

    1. It explains quite well the research ground.
    2. Their computational pipeline is novel;
    3. Their superior performance compared with the existing approaches;
  • 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.

    There are also some identified weakness of the paper:

    1. Given the current commercial 3D mapping systems (Ensite or Navix) are getting better and better for visualizing everything in 3D, detecting catheters in fluoroscopy image were important in the past. But I believe it is less an issue, right? Then how the authors argue the clinical relevant for the proposed study?
    2. Given the limited public data utilised in this study, how the authors can convince us that their models are not overfitted given their approach is so complex (have three components)? that is say, how they can confirm their superior performance is real if we test it on a new dataset?
    3. minor issues, such as RF and CS need to be defined before they are used in the abstract.
  • 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?

    Good to explicitly state the data and codes associated with the study for reproducibility.

  • 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

    There are also some identified weakness of the paper and please address these in the revised manuscript.

    1. Given the current commercial 3D mapping systems (Ensite or Navix) are getting better and better for visualizing everything in 3D, detecting catheters in fluoroscopy image were important in the past. But I believe it is less an issue, right? Then how the authors argue the clinical relevant for the proposed study?
    2. Given the limited public data utilised in this study, how the authors can convince us that their models are not overfitted given their approach is so complex (have three components)? that is say, how they can confirm their superior performance is real if we test it on a new dataset?
    3. minor issues, such as RF and CS need to be defined before they are used in the abstract.
  • 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?

    See my comments for the weakness of the paper.

  • Reviewer confidence

    Somewhat confident (2)

  • [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 proposes a novel method for detecting electrodes in X-ray images, that takes advantage of the a-priori knowledge that the electrodes are designed to be evenly distributed along the catheter. A U-Net like architecture is proposed for this task and the cost function adapted to take into account the spatial-context information.

  • 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 addresses a clinically relevant problem -There is novelty in the architecture as well as the cost function that is adapted to the task at hand -the training and test sets consists of a large number of images, which is multi-center -comparison against SoTA methods and the proposed method is outperforming them

  • 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.

    As is the method is presented no temporal information is taken into account when classifying a new image. Can the authors comment on the temporal consistency of the proposed method? Can the authors provide inference times for the proposed method?

  • 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?

    -data and code are private thus direct reproducibility is difficult -implementing the method based on the details in the paper is challenging

  • 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 has novelty elements and takes into account domain knowledge in order to improve the electrode localization task. The use of SAM to detect the catheter is a nice addition to the method. A number of ablation studies have been performed to investigate the effects of various components in the presented model.

  • 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?

    Well written paper that is of interest to the MICCAI community. Impressive test/training datasets and the proposed method is better that the prior state of the art. Authors propose a novel loss function that is adapted to the task at hand.

  • 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 #3

  • Please describe the contribution of the paper

    This paper presents a landmark detection tool to identify electrodes on a catheter via fluoroscopic images. They use deep learning for spatial-context information. Methods include extracting global and local features using U-Net, catheter segmentation mask using SAM, and a custom weighted loss function.

  • 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 three steps for landmark detection feel dubious at first. The authors justify and demonstrate the significance of each step using an ablation study.
    • SOTA comparisons show superior performance of this study.’
    • Multiple and diverse datasets were used.
    • Methods are well explained and written out.
  • 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.
    • I am unclear about the importance of “highly accurate” landmark detection. Looking at the images, it seems that the detected landmarks (yours or others) are often slightly away from the GT but still on the shaft of the catheter, along its true axis. (Apart from the tip) Does it really matter if the detected-and-GT electrodes are overlapping perfectly, given that the task is to detect the position of the catheter?
  • 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 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?

    N/A

  • 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

    Some suggestions are:

    • Excellent work and great writing.
    • Please discuss the number of landmarks present on a catheter and the images, and their potential variations in clinical practice.
    • Mention the imaging modality i.e. fluoroscopy in the beginning few paragraphs of the introduction. As well as talk about the radiographic appearance and challenges of catheters/electrodes in x-rays.
    • Mention in figure 3 what the numbers respresent - MRE?
  • 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?

    Methods, comparisons with literature, and ablation study supporting the effectiveness of the methods are all well-done.

  • 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



Author Feedback

We are grateful for the time and effort you put into your detailed review. The points you highlighted have been immensely helpful in revising our paper.

Responds to R1: Comment1. Thanks for your comments. In recent years, 3D systems have indeed become increasingly user-friendly, but most doctors still need the assistance of medical imaging during surgery, especially X-rays. The localization of key electrode points not only aids doctors in determining positions but also plays a crucial role in subsequent registration, reconstruction, and other processes. We hope this method can serve as a cornerstone for the future development. Comment2. To avoid overfitting, we set a validation set to periodically evaluate model performance during training and employ early stopping method. Comment3. Thank you so much. We will include the catheter definition in the appendix.

Responds to R3: Weakness: Thank you for your comments. During our evaluation, we calculate the MRE between the landmarks and GT, where the GT is annotated at the electrode center in the images, and the images are extracted from several sequences. This may ensure a certain level of temporal consistency. Our method in this study indeed does not take temporal information into account, which is a question worthy of further exploration, we will consider this issue in our future work. The average inference time for a single image using our method is 26 ms.

Responds to R4: Weakness: In our study, we find that the electrode landmark detection by other methods sometimes deviate significantly from GT, whereas our method mostly yields results that are much closer. Accurate electrode landmark detection is crucial as it forms the foundation for subsequent tasks such as registration, multi-angle reconstruction, and others. Comment1: Thanks for your comments. In this study, RF and CS catheters are two of the most commonly used catheters in clinical surgery. We will further include the catheter definition in the appendix and discuss the number of landmarks and their potential variations in clinical practice. Comment2: Thank you. We will add the relevant descriptions to the introduction according to your suggestions. Comment3: Yes, it’s MRE. We are sorry for the confusion, we will add descriptions to caption of figure 3.



Meta-Review

Meta-review not available, early accepted paper.



back to top