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

Abstract

Effective preoperative planning is crucial for successful cryoablation of liver tumors. However, conventional planning methods rely heavily on clinicians’ experience, which may not always lead to an optimal solution due to the intricate 3D anatomical structures and clinical constraints. Lots of planning methods have been proposed, but lack interactivity between multiple probes and are difficult to adapt to diverse clinical scenarios. To bridge the gap, we present a novel 3D Differentiable-Gaussian-based Planning Strategy (3DGPS) for cryoablation of liver tumor considering both the probe interactivity and several clinical constraints. Especially, the problem is formulated to search the minimal circumscribed tumor ablation region, which is generated by multiple 3D ellipsoids, each from one cryoprobe. These ellipsoids are parameterized by the differentiable Gaussians and optimized mainly within two stages, fitting and circumscribing, with formulated clinical constraints in an end-to-end manner. Quantitative and qualitative experiments on LiTS and in-house datasets verify the effectiveness of 3DGPS.

Links to Paper and Supplementary Materials

Main Paper (Open Access Version): https://papers.miccai.org/miccai-2024/paper/0132_paper.pdf

SharedIt Link: pending

SpringerLink (DOI): pending

Supplementary Material: https://papers.miccai.org/miccai-2024/supp/0132_supp.pdf

Link to the Code Repository

N/A

Link to the Dataset(s)

N/A

BibTex

@InProceedings{Wan_3DGPS_MICCAI2024,
        author = { Wang, Ce and Huang, Xiaoyu and Kong, Yaqing and Li, Qian and Hao, You and Zhou, Xiang},
        title = { { 3DGPS: A 3D Differentiable-Gaussian-based Planning Strategy for Liver Tumor Cryoablation } },
        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

    This paper proposed a 3D differentiable gaussian-based planning strategy for cryoablation of liver tumor considering both the probe interactivity and several clinical constraints.

  • 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.

    1. This paper proposed a parameterized three-dimensional differentiable Gaussian distribution based on the multiple regular circumscribed ellipsoid modeling, which can circumscribe the target tumor.

    2. The proposed 3DGPS builds the relationship between probes when more than one probe is necessary.

  • 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.

    1. The proposed method lacks comparison with other methods, and it is difficult to validate the proposed method to be competitive.

    2. This paper contains insufficient test data, relying on only five cases, which may lead to chance.

    3. The proposed method seems to have no any advantage in planning efficiency, and steps A, B, and C all add up to more than 1 hour.

  • 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?

    It is recommended that the used datasets and code be open-sourced to improve the reproducibility of the proposed method.

  • 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

    1. This paper should add some comparative experiments with other methods to show its competitiveness.

    2. There are many cases in the LiTS dataset, why are only 5 cases selected for testing? More test cases should be added to illustrate the reliability of the method.

    3. The proposed method seems to have no any advantage in planning efficiency, and steps A, B, and C all add up to more than 1 hour.

    4. The paper states that the modified Khachiyan’s algorithm [19] was used. What aspects were improved?

    5. It is recommended that the used datasets and code be open-sourced to improve the reproducibility of the proposed method.

    6. In the paper, the symbols of some formulas and the three metrics (NP, CP, AE) were not explained.

    7. Equation (1) seems to be missing the left part of the equal sign.

  • 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 Reject — could be rejected, dependent on rebuttal (3)

  • Please justify your recommendation. What were the major factors that led you to your overall score for this paper?

    The paper lacks comparative experiments to prove its competitiveness, has too little test data, and requires more than an hour of surgical planning time.

  • 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

    Although this work has limitations in contrast to other methods, surgical planning procedures are frequently difficult to replicate. I believe that the 3D Differentiable Gaussian-based Planning Strategy (3DGPS) for cryoablation of liver tumors, which takes into account both probe interactivity and a number of clinical restrictions, is a innovative surgical planning method with potential clinical applications.



Review #2

  • Please describe the contribution of the paper

    Effective preoperative planning is vital for successful liver tumor cryoablation. Traditional methods rely on clinician experience and struggle with complex 3D structures and clinical constraints. This paper propose a novel 3D Differentiable Gaussian-based Planning Strategy (3DGPS) is proposed to address the aforementioned issue. It considers probe interaction and clinical constraints, aiming to minimize the ablation region. Quantitative and qualitative experiments confirm its effectiveness.

  • 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.

    Formulating the planning problem as modeling multiple regular ellipsoids and parameterizing these ellipsoids using differentiable 3D Gaussians is a technically innovative and mathematically solution. The experimental evaluation on real-world datasets provides strong empirical validation of the proposed method’s efficacy.

  • 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.

    Discuss the potential impact and broader applicability of the proposed method, beyond just liver tumor cryoablation, if applicable.

  • Please rate the clarity and organization of this paper

    Excellent

  • 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

    See the strenghts and weaknesses above.

  • 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 the strenghts and weaknesses above.

  • Reviewer confidence

    Not confident (1)

  • [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

    1.Formulate the probe insertion problem as the multiple regular circumscribed ellipsoid modeling, to circumscribe the target tumor. 2.Build the relationship between probes, which is better to provide the optimal number choice and the trajectory information, along with alternative suboptimal plans.

  • 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 tumor cryoablation area is parameterized by the differentiable Gaussians and optimized mainly within two stages, fitting and circumscribing, with formulated clinical constraints in an end to end manner.

  • 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.

    Should do more experiments to compare the proposed method with other related methods that used for ablation operation, As I know, there are some works about wave ablation, which has the similar strategy with this paper.

  • 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?

    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

    1.For the tumor cryoablation, the vessels around the tumor will affect the operation. In the operation planning, the vessels’ location should be considered.

    1. More experiments for comparison should be listed.
  • 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?

    Formulate the probe insertion problem as the multiple regular circumscribed ellipsoid modeling, to circumscribe the target tumor. Build the relationship between probes, which is better to provide the optimal number choice and the trajectory information, along with alternative suboptimal plans.

  • 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 Reject — could be rejected, dependent on rebuttal (3)

  • [Post rebuttal] Please justify your decision

    In the rebuttal, the authors did not make the problems clear.



Review #4

  • Please describe the contribution of the paper

    This work proposes a 3D Gaussian-based planning approach for liver tumor cryoablation. Results from simulation experiments using a public dataset and a single-case clinical experiment (compared with the actual ablation zone), demonstrate the feasibility of the proposed approach. Overall, this work holds its clinical value.

  • 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 and organized, effectively delivering its content to readers.

    This work demonstrates its clinical value by providing multiple potential planning trajectories for doctors, and shows promise for benefiting other procedures, including liver thermal ablations. And the methodology of this planning strategy is interesting.

  • 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.

    For the proposed method, it might be challenging for readers to re-implement it, because certain details (e.g. equation (2)) are not clearly described. It would be beneficial for the authors to make the code or provide more detailed information publicly available.

    To demonstrate the effectiveness of this planning strategy, one important aspect is that whether the estimated tumor coverage from the planning approach can be successfully applied to actual clinical cases. Otherwise, the residual tumor may occur. However, in this work, the reviewer did not find results/discussion to consider this point.

  • 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?

    It would be beneficial for the community to make the code publicly available.

  • 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

    In a single-case clinical experiment, please clarify whether the treatment was conducted by Plan C. If so, which specific plan (#P) is actually used?

    The concept of “probe interactivity” needs further clarification. In the context of ablation procedures, the resulting ablation zone from multiple probes typically does not simply aggregate the individual zones created by each probe. The interaction between probes/ablation zones is complex, associated with patient tissue properties. However, in this work, the term “probe interactivity” appears to introduce a different definition, please clarify that to avoid the confusion.

    In page 2, please use the full name of “KA” algorithm.

  • 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?

    This work demonstrates its clinical value and shows promise for benefiting other procedures, including liver thermal ablations.

  • 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

    The reviewer appreciates the responses, but one major comment “One important aspect is that whether the estimated tumor coverage from the planning approach can be successfully applied to actual clinical cases.” seems not to be answered. The reviewer decided to keep the previous decision.



Author Feedback

We sincerely thank AC and all reviewers for their time and efforts to give useful comments. We below clarify the main concerns (C) first, and then reply to other questions (Q). We will revise the paper carefully to address all comments.

C: (@R1\&R4) Lacking comparison with other methods.

A: In fact, we are eager to compare with existing works, but they do not release the source codes. Moreover, we observe that they only compared between their proposed different versions [3, 11, 13, 15]. Besides, due to the properties of tumors, AE values vary greatly in different cases. Since they do not provide the specific testing CASE IDs, we could only compare our plans A-C.

Particularly, compared coarsely with [13] on LiTS, where their average AE is 60.02, our AE result (62.53) of Plan C is better. Note again that the comparison is really meaningless as the case IDs do not match since their case IDs are unknown. Further, in terms of the methodology, we have provided an advantageous analysis of our method in the introduction, compared with recent methods.

Q1: (@R1) The planning efficiency problem.

A: The proposed Plans A-C are independent and do not require the sum of computation times. The computation time depends on the case complexity. Particularly Plan C takes 21-43 minutes, which is comparable to recent optimization-based methods [11, 13]. Besides, the efficiency of our method has considerable room for improvement.

Q2: (@R1) Insufficient test data.

A: Our current version focuses on radical tumor ablation, which clinically requires the tumor to be smaller than 50mm and limits our choice of LiTS. Besides, we have referred to several papers [11, 13, 15], and they also conduct experiments on a limited number of 10-15 patients. Moreover, since our method is optimization-based and independent of tumor shape, the result is reliable. We will show the results of more cases along with our code.

Q3: (@R3) The potential impact and broader applicability.

A: Except for the liver tumor cryoablation, our method can be directly generalized to kidney and other tumor ablation. Further, our method is modality-agnostic and can be applied to other modalities, e.g. MRI.

Q4: (@R4) The vessels’ location should be considered.

A: Our method has considered the conflict between ablation zones and vessels via segmentation. Specifically, we integrate vessel masks in L_overlay and L_probe, avoiding vessels existing in the ablation zones and probe trajectories.

Besides, our current version is based on TotalSegmentator [21], which can segment some liver vessels and can be improved via clinical doctors to include more accurate vessels that potentially affect the operation.

Q5: (@R5) Clarifying whether the single-case treatment was conducted by Plan C.

A: The treatment was conducted empirically by clinical doctors, not based on Plan C. Comparisons on the IH dataset are performed to validate our plans. Especially the metric IoU quantitatively indicates the overlaps between empirical zones and our optimized zones.

Q6: (@R5) The concept of “probe interactivity” needs clarification.

A: Thanks for clarifying the definition of probe interactivity, which is exactly what we want to express.

In our method, we intend to model the probe/ablation zones with parameterized Gaussians/ellipsoids. In this manner, the interaction between ellipsoids represents the corresponding probe interactivity, which can be optimized automatically with our method.

Q7: (@R5) The reproducibility of the method.

A: We have tried our best to give the implementation details for reproducing. To understand Eq. (2) and other equations clearly, we have clarified the meaning of each variable following each equation and provided the values of thresholds we used in Section 3.
And, we will release our code for usage later.

Besides, we will accordingly modify our paper, e.g. giving the full name Khachiyan’s Algorithm (KA) on Page 2. (@R5)



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.

    Reject

  • Please justify your recommendation. You may optionally write justifications for ‘accepts’, but are expected to write a justification for ‘rejects’

    Many thanks to the rebuttal from the authors. My main concern is that the issues from R#4 are not fully addressed. This paper is quite borderline.

  • 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).

    Many thanks to the rebuttal from the authors. My main concern is that the issues from R#4 are not fully addressed. This paper is quite borderline.



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’

    The rebuttal has adequately addressed the major issues.

  • 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).

    The rebuttal has adequately addressed the major issues.



Meta-review #3

  • 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’

    There is interest in this work among reviewers and the rebuttal reasonably addresses the critiques.

  • 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).

    There is interest in this work among reviewers and the rebuttal reasonably addresses the critiques.



back to top