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

Authors

Yiran Wei, Chao Li, Stephen John Price

Abstract

Brain tumors are characterised by infiltration along the white matter tracts, posing significant challenges to precise treatment. Mounting evidence shows that an infiltrative tumor can interfere with the brain network diffusely. Therefore, quantifying structural connectivity has potential to identify tumor invasion and stratify patients more accurately. The tract-based statistics (TBSS) is widely used to measure the white matter integrity. This voxel-wise method, however, cannot directly quantify the connectivity of brain regions. Tractography is a fiber tracking approach, which has been widely used to quantify brain connectivity. However, the performance of tractography on the brain with tumors is biased by the tumor mass effect. A robust method of quantifying the structural connectivity in brain tumor patients is still lacking.

Here we propose a method which could provide robust estimation of tract strength for brain tumor patients. Specifically, we firstly construct an unbiased tract template in healthy subjects using tractography. The voxel projection procedure of TBSS is employed to quantify the tract connectivity in patients, based on the location of each tract fiber from the template. To further improve the standard TBSS, we propose an approach of iterative projection of tract voxels, under the guidance of tract orientation measured by voxel-wise eigenvectors. Compared to the conventional tractography methods, our approach is more sensitive in reflecting functional relevance. Further, the different extent of network disruption revealed by our approach correspond to the clinical prior knowledge of tumor histology. The proposed method could provide a robust estimation of the structural connectivity for brain tumor patients.

Link to paper

DOI: https://doi.org/10.1007/978-3-030-87234-2_49

SharedIt: https://rdcu.be/cyl8L

Link to the code repository

https://github.com/vctorwei/TumorBrainConnectivity

Link to the dataset(s)

N/A

Reviews

Review #1

Please describe the contribution of the paper

In this manuscript, the authors propose a method to estimate connectivity based on the construction of a structural network in patients with brain tumor (glioma, glioblastoma and meningioma). The method is based on TBSS and on a tractography template from healthy controls. A simple but interesting clinical validation is presented using graph-theory.
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 manuscript is clearly written and introduces all the necessary concepts. The figures are mostly clear and compelling. The topic addressed by the proposed method has valuable impact in clinical application for brain tumor patients. The set of experiments is convincing and involves three different cohorts of subjects. The method has a preliminary clinical validation based on the graph-theory concept of network efficiency.
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 is no real algorithmic contribution in this manuscript. The novelty is in using known building blocks in an interesting way. The comparison with TBSS is not particularly interesting because TBSS in a known underpowered method. With respect to the usual technical contribution expected in MICCAI submissions, the one presented in this manuscript can be considerate as “moderate”.
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

The method is clearly explained but the code used in the experiments is not made available. The details of which specific subjects from public datasets were used in the study are not reported. With substantial effort, the paper could be reproduced on analogous data, expecting comparable results. But it is not possible to repeat the exact steps presented in the manuscript.
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 technical details of the proposed method should be described more. Some steps can only be understood from the figures, which is not how it should be.
- The statistical tests report p-vaules in a questionable way: always as “p<0.01” or “p<0.05” etc. The authors should report the actual value of p and not under which threshold they find appropriate to present it: they do a Fisherian significance test and not a Neyman-Pearson hypothesis test! Moreover, it is meaningless to use different thresholds across the manuscript. To conclude, the authors should consider 0.005 or less as an acceptable threshold for significance uniformly across the manuscript (see Banjamin et al. 2017 Nature Human Behavior).
- There are minor typos here and there: Sec. 1.4, last word (“Fig. 1”, which has no context), Sec. 2.2 p.5 (“principle” should be “principal”), Sec. 2.6 (“Clincial” should be “Clinical”), Sec. 3.4 (“significantly” should be “significant”).
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?
- The manuscript is well written.
- The application is appealing.
- The experiments looks convincing.
- The comparison against TBSS is not particularly motivating.
- The technical contribution is moderate.
What is the ranking of this paper in your review stack?

1
Number of papers in your stack

4
Reviewer confidence

Confident but not absolutely certain

Review #2

Please describe the contribution of the paper

The authors proposed a method to study the structural connectivity in brain tumor patients. The idea is to construct a healthy control template and then quantify patient specific connectivity by comparing the template. The results are compared to fMRI data, showing a high structural-functional correlation.
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.

. Overall, this is a well-designed study and provides a quite interesting clinical application.
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 have some minor comments:
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

Code not mentioned in the paper.
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 authors proposed a method to study the structural connectivity in brain tumor patients. The idea is to construct a healthy control template and then quantify patient specific connectivity by comparing the template. The results are compared to fMRI data, showing a high structural-functional correlation. Overall, this is a well-designed study and provides a quite interesting clinical application. I have some minor comments:
1. A more comprehensive literature review about using tractography in tumor patient studies should be done. In the current manuscript, Section 1.1 does not have any references and Section 1.2 has only a few. Now, it is hard for a reader to understand that this is an important task.
2. Page 3. UKF is not designed for brain tumor patients. I think it is designed for general tractography purpose, and is shown to be effective in tumor patients.
3. Section 2.1. please give the exact number of patients in each dataset, and provide acquisition details.
4. Section 2.3 The implementation of UKF is provided in SlicerDMRI, not just 3D Slicer.
5. Section 2.3 Please clarify how the mean FA is computed.
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?

interesting topic and well designed method
What is the ranking of this paper in your review stack?

1
Number of papers in your stack

3
Reviewer confidence

Confident but not absolutely certain

Review #3

Please describe the contribution of the paper

The authors proposed a new method for solving the tumor mass effect in estimating the structural connectivity (SC) (i.e., TBSS) in patients with brain tumor. The authors clarified the rationale and its clinical importance of the new method. The novel section was to use the voxel-based eigenvector to bypass the discontinuity and the enhanced performance was reflected on the coupling between SC and functional connectivity (FC), as well as the global efficiency. Although the proposed method seemed technical sound, the supporting indices were less effective for justification. However, the estimation target of SC-FC coupling may be misleading because the FC is derived from the resting-state fMRI. Currently, the SC-FC coupling is inconsistent even in the normal participants, unless the participants is in sleeping. It means that the SC-FC coupling on the brain-tumor patient may not be a valid justification. Besides, the large disparity between the proposed method and UKF remains to be unexp
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 proposed method is reasonable and useful in clinical practice, especially on the presurgical mapping.
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.

Using the global efficiency as a justification index may be inadequate since the global efficiency does not provide underlying physiological mechanisms. Therefore, it is suggested to present the individual FA skeleton with a 3D viewpoint following the protocol of presurgical mapping for justification.
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

Establishing a rapid and reasonable method is good in clinical practice.
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

Using the global efficiency as a justification index may be inadequate since the global efficiency does not provide underlying physiological mechanisms. Therefore, it is suggested to present the individual FA skeleton with a 3D viewpoint following the protocol of presurgical mapping for justification.
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?

Establishing a rapid and reasonable method is good in clinical practice. The authors proposed a new method for solving the tumor mass effect in estimating the structural connectivity (SC) (i.e., TBSS) in patients with brain tumor. The authors clarified the rationale and its clinical importance of the new method. The novel section was to use the voxel-based eigenvector to bypass the discontinuity and the enhanced performance was reflected on the coupling between SC and functional connectivity (FC), as well as the global efficiency.
What is the ranking of this paper in your review stack?

4
Number of papers in your stack

4
Reviewer confidence

Confident but not absolutely certain

Review #4

Please describe the contribution of the paper

This paper investigates an analytic approach for understanding changes in connectivity that occur due to brain tumors. The proposes a diffusion MRI analysis pipeline and compares it with standard tractography approaches.
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 problem of measuring connectivity strength in brain tumor cases is an important one, and an area where more attention is warranted. The comparison with standard tractography methods was a good part of the evaluation, as well as the circle visualization of subnetworks.
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 think there are many important technical details which are not sufficiently described by the paper, making it difficult to understand and properly evaluate it. For example, in the joint structural and functional MRI analysis, there are many important modeling choices that need to be explained, as they can have dramatic effects on the results. I think the other major limitation of the work is that the accuracy of the template registration may not sufficiently align the template-defined connections with the subject data. They are likely okay in deeper white matter structures, but deformable registration of diffusion datasets often has issues with the precise alignment superficial white matter and cortex features. One way to address this would be to compare the atlas-defined tracts with subject-space-defined tracts (from diffusion models in the subject data) to see how agree quantitatively.
Please rate the clarity and organization of this paper

Poor
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

There are some important details of the analysis which are not described in sufficient detail to be reproduced.
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

Diffusion modeling approach is just as important as tractography reconstruction technique. For example, multi-tensor modeling can possibly help address the challenges with edema, e.g. using the ball-and-sticks model or free-water-eliminated DTI. FACT and UKF are reasonable approaches to compare against, but I wouldn’t consider them state-of-the-art.
Please state your overall opinion of the paper

probably reject (4)
Please justify your recommendation. What were the major factors that led you to your overall score for this paper?

I think the paper has good potential, but there were major issues with the clarity in describing the methods and technical issues that may need additional validation experiments.
What is the ranking of this paper in your review stack?

4
Number of papers in your stack

4
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.

The authors propose to improve tractography-based computation of brain structural connectivity from diffusion-MRI scans of tumor patients using a template from healthy subjects. Most reviewers agree that the paper should be accepted. I think the criticism of the work, namely insufficient detail and comparison with the state of the art, some statistical concerns, and non-ideal evaluation metrics, should not preclude the paper from MICCAI.
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).

3

Author Feedback

We appreciate all the reviewers for their insightful comments. We have expanded the literature review to introduce the clinical significance of the research topic. We also added more details to the description of the method and testing datasets, corrected the statistical concerns regarding the p values and checked the typos/inappropriate descriptions throughout the paper. Code will be available on GitHub.

We agree with the comments of the Reviewer 3 about the SC-FC coupling and global efficiency. Although similar methods are used by other studies in the lesioned brain (H Chen, 2021), these two metrics have limitations. Therefore, we have used tumor histology/malignancy as a strong evidence to support the proposed method. Our results are in line with the prior knowledge that three tumor types have different extent of white matter disruption. Due to the limitation of data availability, we cannot add more metrics for validation. Further investigation using more clinical measures are warranted.

Regarding the comments of reviewer 4, we would like to clarify that our method leverages a tract atlas generated from healthy controls, therefore the deformable co-registration is essential for mitigating the mass effect of the tumor. In addition, several studies showed that deformable registration is feasible for both tract template (A Salvalaggio, 2020) and the brain with tumor (P Prasanna, 2019). We have compared this method with the tractography methods that are performed at the patient-specific space. Our results demonstrate the superiority of the proposed method. The additional comparison with diffusion modelling approach is appealing in future work.

back to top

Quantifying structural connectivity in brain tumor patients