Paper Info Reviews Meta-Review Author Feedback Post-rebuttal Meta-Reviews

Authors

Lixin Ren, Shang Wan, Yi Wei, Xiaowei He, Bin Song, Enhua Wu

Abstract

Portal hypertension is one of the major complications in patients with chronic liver diseases (CLD) which induces an increase in portal vein gradient pressure. At advanced stages, it can cause the esophageal varices and variceal hemorrhage. Therefore, portal hypertension has been the leading cause of mortality in CLD patients. To diagnose portal hypertension, the invasive hepatic venous pressure gradient (HVPG) measurement is still the only validated technique to accurately evaluate changes in portal pressure and regarded as the standard reference. However, it entails the limitation of invasive procedure and have the risk of further bleeding and inflammation. In this paper we propose an Eulerian computational fluid dynamics (CFD) model to facilitate hemodynamics analysis. To enable consistent simulation results with different boundary conditions, a diffuse boundary handling technique was proposed to impose smooth boundary conditions for both the pressure and velocity fields. We also propose a computational workflow for quantifying patient-specific hemodynamics in portal vein systems non-invasively. The simulation is performed on patient-specific PV models reconstructed from CT angiographic images. Experiments show that pressure changes in the PV of patients with portal hypertension due to blockage of the RPV is significantly lower than that of normal subjects.

Link to paper

DOI: https://doi.org/10.1007/978-3-030-87240-3_11

SharedIt: https://rdcu.be/cyl5H

Link to the code repository

N/A

Link to the dataset(s)

N/A

Reviews

Review #1

Please describe the contribution of the paper

The authors propose a novel non-FEM CFD approach to estimate pressure in the portal vein system, which may have clinical application in liver disease.
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 very well written and easy to read. The figures are clear and appealing. The proposed clinical application is interesting and the proposed method has some interesting features.
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 study appears to be very preliminary, comparing the proposed methods to FEM in a very idealised situation. The method does not appear well validated in patients, showing very large discrepancies when compared to HVPG, which are not addressed in the Discussion. The manuscript should include a more detailed literature review, in my opinion. It is not clear from it the CFD SOTA for the portal system is and therefore how the proposed method compares to it and may have clinical advantages.
Please rate the clarity and organization of this paper

Very Good
Please comment on the reproducibility of the paper. Note, that authors have filled out a reproducibility checklist upon submission. Please be aware that authors are not required to meet all criteria on the checklist - for instance, providing code and data is a plus, but not a requirement for acceptance

This is an exploratory study, which is not suitable for a large reproducibility analysis. Such an analysis is not mentioned in the manuscript. It is not clear whether the code and/or the used dataset are made 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: https://miccai2021.org/en/REVIEWER-GUIDELINES.html

The Introduction would be improved by referring to previous CFD studies of the portal vein system, instead of focusing so much on the details of the pathology. Likewise, the methods are written as if the authors are proposing a new formalism for CFD rather than relying on existing established techniques (such as finite differences methods) and approximations. For example: “For simplicity, we assume the blood behaves as an incompressible Newtonian fluid and the vessel walls are rigid.” -> Please discuss in which conditions these assumptions may not hold/data that shows these assumptions are typically valid. Also, for the experiments, it is not clear what numerical the implementation details are. What was the spatial size of the grid, for example? How sensitive to the numerical choices are the results (e.g. to the inlet flat (!) velocity profile of 0.1364 m/s)? How do the simulations compare to FEM in terms of run time, etc? From Fig 5, there is little difference between the anatomy of the PV in healthy subjects and patients. What explains the difference in Dp between the two groups then? Were different system parameters used for this? The Conclusions should explicitly describe the advantages of the proposed method relative to existing ones and the potential impact of this method.
Please state your overall opinion of the paper

reject (3)
Please justify your recommendation. What were the major factors that led you to your overall score for this paper?

Unfortunately it was not clear to me what the advantages of the proposed method are relative to existing ones. The validation data did not appear to support the conclusion that this method was accurate and I did not fully understand the details of the proposed method.
What is the ranking of this paper in your review stack?

3
Number of papers in your stack

3
Reviewer confidence

Confident but not absolutely certain

Review #2

Please describe the contribution of the paper

In this paper, an Eulerian Computational Fluid Dynamics (CFD) model is proposed to quantify patient-specific hemodynamics in portal vein system non-invasively. A diffuse boundary condition approach is proposed to mitigate the effect of the choice of boundary conditions for the pressure and velocity fields. Experiments on a synthetic case where analytical solution is known (Poiseuille flow) shows that the proposed approach is indeed less sensitive to the imposed boundary condition. The model is also evaluated on 6 patient-specific portal vein geometries. The simulated pressure gradient seems to correlate well with invasive hepatic venous pressure gradient measurements for the 3 patients. The method is able to differentiate between the 3 patients with portal hypertension and normal subjects.
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.

Using a CFD model to estimate portal hypertension in order to avoid using invasive measurements and its related complication is very interesting and of clinical importance. The modeling hypothesis proposed in the paper: the blockage of the right portal vein will show a high difference in portal vein pressure for normal subject compared to patient suffering from portal hypertension since the portal vein system already compensates for that. This hypothesis is sound and supported by results on 6 patients. The idea to use diffuse boundary condition to be less sensitive to the boundary condition choice is interesting and validated on a simplified set-up. The use of a distance-signed function to limit the stair-step grid artifacts resulting from the creation of the computational domain from a binary mask is also nicely presented. Finally, the results on 6 patients are promising.
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.

In the Poiseuille flow experiment, only one value of constant flat profile is tested. How are the results affected by this value? It would have been interesting to see results for different values.

Some implementation details are not given in the method:

How is the mesh generated? Using which tool? What is the mesh resolution? How is the centerline extracted?

For the in-vivo patient experiment: How is the portal vein segmentation done? Manually? From which CT phases? I assume from the portal one but this is not clear. What are the Boundary conditions imposed at the outlets ? What is the spatial resolution used?

In the Poiseuille flow experiment, what are the values of L, mu ?
Please rate the clarity and organization of this paper

Good
Please comment on the reproducibility of the paper. Note, that authors have filled out a reproducibility checklist upon submission. Please be aware that authors are not required to meet all criteria on the checklist - for instance, providing code and data is a plus, but not a requirement for acceptance

Some numerical implementation details are specified.
Please provide detailed and constructive comments for the authors. Please also refer to our Reviewer’s guide on what makes a good review: https://miccai2021.org/en/REVIEWER-GUIDELINES.html

Computational time in the in-silico and in-vivo experiments as well as Hardware specification would be interesting to report.

In the case discussed in Fig 2, since u_i is inside the vessels, It is not clear why Phi_i is in a range of [-d,d] and not [0,d]

It is not clear how the author come up with Eq (8) computing the divergence of the velocity.

It should be made clear that no HVPG are available for the 3 normal subjects.

What is the CT reconstruction thickness: 1 or 2 mm?

There is a sentence repetited in the abstract.

Some typo are present in the text. A careful proof-reading would benefit the paper.
Please state your overall opinion of the paper

Probably accept (7)
Please justify your recommendation. What were the major factors that led you to your overall score for this paper?

Interesting use of CFD model to get rid of invasive diagnostic measurement. Interesting modeling hypotheses tested and validated on 6 patients Interesting handling of the boundary condition which is a well-known issue in hemodynamics modeling.
What is the ranking of this paper in your review stack?

2
Number of papers in your stack

4
Reviewer confidence

Very confident

Review #3

Please describe the contribution of the paper

The authors tackle the modeling of hepatic, portal vein hypertension through derived (geometry) CFD. Clinical measure of portal vein pressure gradients is complex, error prone and invasive. The image-based, CFD methods developed here can create pressure predictions from non-invasive, cross sectional (CT) image data. One key contribution by the authors addresses discontinuities possible when boundaries conditions on compute cells violate geometry constraints. The boundary condition handling is developed so as to generate reduced error. Simulation results are presented and limited human data.
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.

Strengths of the paper include: well structured; explanation of methods sufficient and not overly-detailed, an important clinical prognostic value can be generated by this technique and would be a powerful, imaging biomarker, and the numerical considerations of BC handling seems novel/unique,
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 gaps in writing clarity that make the interpretation difficult at times. There is a lack of human-geometry with reference standards so the ultimate performance (at this point) is somewhat hard to gauge - although I’m sure the analysis is ongoing.
Please rate the clarity and organization of this paper

Good
Please comment on the reproducibility of the paper. Note, that authors have filled out a reproducibility checklist upon submission. Please be aware that authors are not required to meet all criteria on the checklist - for instance, providing code and data is a plus, but not a requirement for acceptance

Reproducibility would be challenging, but not impossible. The contribution of meshing code, simulation datasets and data would be very useful.
Please provide detailed and constructive comments for the authors. Please also refer to our Reviewer’s guide on what makes a good review: https://miccai2021.org/en/REVIEWER-GUIDELINES.html

Page 2 - do you assume full blockage of outlets? Important to note. Page 3 - The presentation raises some questions - do you mean cells as in “blood cells” or a “cell” in the geometry? I don’t think you are trying to model individual or groups of blood cells, but rather compute/predict the velocity fields so as to calculate shear/strain on a blood cell at a certain location. From the earlier description, isn’t the geometry a “tube” with fluid domain vs wall? Are you trying to model different levels of wall material? I don’t understand this sentence, then. What is the mask applied too? Page 5 - you state that “all patients underwent CT angiography”. Now you commnet “normal subjects only underwent CTA”. Where there abnormal subjects in the cohort - and if so, what classifies normal vs abnormal? Page 6 - Did you enforce heuristics to how much of geometry to prune after the outlet? Likewise for the inlet condition. Page 8 - How do these compare with the clinical findings (pressure difference)? How many patients were actually enrolled and analyzed? Some presentation of a trending would be useful. Its probably an issue of timing - the data sets were just being collected etc.. it’s hard, however, to judge the true performance and utility of the approach without more comparison data. The in-vivo results are encouraging and there is great utility in not being dependent on specific assumptions of flow profile and type. Further validation, however, is warranted.
Please state your overall opinion of the paper

accept (8)
Please justify your recommendation. What were the major factors that led you to your overall score for this paper?

The methods are sound, the application is important, and there is value in discussion/feedback from peers in open-session and in the proceedings.
What is the ranking of this paper in your review stack?

3
Number of papers in your stack

5
Reviewer confidence

Confident but not absolutely certain

Primary Meta-Review

Please provide your assessment of this work, taking into account all reviews. Summarize the key strengths and weaknesses of the paper and justify your recommendation. In case you deviate from the reviewers’ recommendations, explain in detail the reasons why. In case of an invitation for rebuttal, clarify which points are important to address in the rebuttal.

This paper presents a potentially interesting contribution to the field of computational hemodynamics. The proposed methodology tackles some of the ever lasting problems of personalised hemodynamics problems (sensitivity to BCs, identification of locations of high pressure, definition of geometry, etc.). The manuscript is clear and well written. Reviewers raised that discussion could be improved in some aspects and that how the proposed method differentiates from the SOTA. Details of the numerical implementation could be improved, this was explicitly raised by to reviewers.
What is the ranking of this paper in your stack? Use a number between 1 (best paper in your stack) and n (worst paper in your stack of n papers).

4

Author Feedback

We thank the reviewers for their thoughtful comments and suggestions. Following are answers to major points that reviewers concern.

Contribution of our work and how our method can be distinguished from the SOTA. (#Reviewer 1) Answer: The long-term objective of our work is to find an CFD-based identifier that can help diagnose portal hypertension noninvasively. However, previous works on how to address this problem are rather limited. Among those few works, most of them only apply the FEM solver to solve hemodynamics in the PV system, but pay no attention on the numerical problems (please refer to a recent work by Qi et al.[2018]). However, as demonstrated with our experiments in Fig.4, the accuracy of a CFD solver is largely influenced by the spatial resolutions, boundary handing techniques and numerical solvers. In order to reduce the sensibility, we propose a diffuse boundary handling technique for an Eulerian CFD model, thus being able to impose smooth boundary conditions for both the pressure and velocity fields. Experiments show that our approach is less sensitive to boundary conditions and is able to reproduce more consistent simulation results under two commonly used BCs.

Reproducibility of our work. (#Reviewer 1, #Reviewer 2, #Reviewer 3) Answer: Due to the paper length limitation, the standard parts of our CFD solver are omitted, e.g., how to solve the linear system of equations. If more details are helpful, we can provide a supplement to show all necessary details. In fact, we are developing an open-source project. We assure the reviews all source codes will be made publicly available in the near future. However, making the dataset publicly available has not yet been approved by our hospital, and we will do more effects to work this through. Besides, the current data size is still small and we are enrolling more patients and volunteers to enrich the dataset.

Below are our answers to other points that we think each reviewer may concern.

#Reviewer 1

Why do the results of in-vivo validation show discrepancies compared to HVPG?
Answer: In general, the invasive HVPG is measured in the right hepatic vein. In our in-vivo validation, the pressure difference is computed in the portal vein. The results are not exactly equivalent, but we can notice a strong correlation as shown in Fig. 5.

How to explain the difference in pressure difference between the two groups in Fig. 5. Answer: The same boundary conditions and parameters are applied to both the patients and the normal subjects. At advanced stages of liver cirrhosis, patients with terminal hepatic failure (THF) were usually observed with atrophy in the right liver lobe while hypertrophy in the left liver lobe [16]. Therefore, our assumption is if the same inflow boundary condition is applied and a blockage is applied on right portal vein, the pressure change in PV systems of patients with portal hypertension is expected to be smaller than those without portal hypertension. Our assumption has also been validated in Fig.5.

#Reviewer 2

In the Poiseuille flow experiment, how are the results affected by inlet flat velocity? Answer: According to our in silico study on a steady Poiseuille flow, the simulation results are not sensitive to different values of inlet speed.

#Reviewer 3

Do you assume full blockage of outlets? Answer: Only the RPV is fully blocked. Please refer to page 6 “In the second simulation, we virtually blocked the outlet of RPV…”.

We thank the reviewers for other detailed and valuable suggestions on how to improve the paper writing. We will fix those problems in the revision.

Post-rebuttal Meta-Reviews

Meta-review # 1 (Primary)

Please provide your assessment of the paper taking all information into account, including rebuttal. Highlight the key strengths and weaknesses of the paper, clarify how you reconciled contrasting review comments and scores, indicate if concerns were successfully addressed in the rebuttal, and provide a clear justification of your decision. If you disagree with some of the (meta)reviewer statements, you can indicate so in your meta-review. Please make sure that the authors, program chairs, and the public can understand the reason for your decision.

The rebuttal to the reviewers and AC comments are correct and highlight the value of the manuscript. This is a valid contribution to the conference, extending the SOTA and tackling some problems in an innovative and distinct manner. The authors committed to providing additional details (as supplementary material) to improve the reproducibility of their method.
After you have reviewed the rebuttal, please provide your final rating based on all reviews and the authors’ rebuttal.

Accept
What is the rank of this paper among all your rebuttal papers? Use a number between 1 (best paper in your stack) and n (worst paper in your stack of n papers).

7

Meta-review #2

Please provide your assessment of the paper taking all information into account, including rebuttal. Highlight the key strengths and weaknesses of the paper, clarify how you reconciled contrasting review comments and scores, indicate if concerns were successfully addressed in the rebuttal, and provide a clear justification of your decision. If you disagree with some of the (meta)reviewer statements, you can indicate so in your meta-review. Please make sure that the authors, program chairs, and the public can understand the reason for your decision.

Reviewers did provide more details on open questions in their rebuttal. For this reviewer, enthusiasm of this paper is still limited by questions related to the presentation, clarity of writing, questions on reproducibility and the somewhat preliminary and exploratory nature of the subject. It seems this paper may not be fully ready for MICCAI. Some major questions of reviewers remain.
After you have reviewed the rebuttal, please provide your final rating based on all reviews and the authors’ rebuttal.

Reject
What is the rank of this paper among all your rebuttal papers? Use a number between 1 (best paper in your stack) and n (worst paper in your stack of n papers).

10

Meta-review #3

Please provide your assessment of the paper taking all information into account, including rebuttal. Highlight the key strengths and weaknesses of the paper, clarify how you reconciled contrasting review comments and scores, indicate if concerns were successfully addressed in the rebuttal, and provide a clear justification of your decision. If you disagree with some of the (meta)reviewer statements, you can indicate so in your meta-review. Please make sure that the authors, program chairs, and the public can understand the reason for your decision.
This paper describes a CFD model to facilitate hemodynamics analysis for chronic liver diseases - more concretely portal hypertension. The reviews were mixed (3,7,8). The main criticism from the negative review was the lack of evidence how the method is different from SOTA.

The strengths identified included
- the method avoids an alternative invasive measurement ( = relevant application)
- sound hypothesis, confirmed on 6 patients
- interesting new methods elements (eg the diffuse boundary condition)
The authors answer the raised major concern in some detail in the rebuttal including differences and shortfalls of current methods. The paper describes an early stage of a new method which can potentially have interesting applications - I would thus recommend accepting the paper. The authors promise to make the code available and hint that once the hospital allows this will also make data available - which will be of benefit for the community.
After you have reviewed the rebuttal, please provide your final rating based on all reviews and the authors’ rebuttal.

Accept
What is the rank of this paper among all your rebuttal papers? Use a number between 1 (best paper in your stack) and n (worst paper in your stack of n papers).

6

back to top

Towards a non-invasive diagnosis of portal hypertension based on an Eulerian CFD model with diffuse boundary conditions