Paper Info Reviews Meta-Review Author Feedback Post-rebuttal Meta-Reviews Back to top
List of Papers By topics Author List
Paper Info Reviews Meta-review Author Feedback Post-Rebuttal Meta-reviews

Authors

Ye Wu, Xiaoming Liu, Xinyuan Zhang, Khoi Minh Huynh, Sahar Ahmad, Pew-Thian Yap

Abstract

Brain tissue microarchitecture is characterized by heterogeneous degrees of diffusivity and rates of transverse relaxation. Unlike standard diffusion MRI with a single echo time (TE), which provides information primarily on diffusivity, relaxation-diffusion MRI involves multiple TEs and multiple diffusion-weighting strengths for probing tissue-specific coupling between relaxation and diffusivity. Here, we introduce a relaxation-diffusion model that characterizes tissue apparent relaxation coefficients for a spectrum of diffusion length scales and at the same time factors out the effects of intra-voxel orientation heterogeneity. We examined the model with an in vivo dataset, acquired using a clinical scanner, involving different health conditions. Experimental results indicate that our model caters to heterogeneous tissue microstructure and can distinguish fiber bundles with similar diffusivities but different relaxation rates. Code with sample data is available at https://github.com/dryewu/RDSI.

Link to paper

DOI: https://doi.org/10.1007/978-3-031-43993-3_15

SharedIt: https://rdcu.be/dnwNg

Link to the code repository

https://github.com/dryewu/RDSI

Link to the dataset(s)

N/A

Reviews

Review #1

  • Please describe the contribution of the paper

    This study presents a novel Relaxation-Diffusion Spectrum Imaging (RDSI) framework that combines multi-TE dMRI and spherical deconvolution to disentangle relaxation and diffusion processes in multi-compartmental tissues. RDSI demonstrates potential in improving contrast and characterization of microstructural changes in gliomas, offering potentially valuable information for clinical assessment of glioma progression and treatment response.

  • 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 Relaxation-Diffusion Spectrum Imaging (RDSI) framework effectively disentangles relaxation and diffusion processes in multi-compartmental tissues, providing a more comprehensive understanding of tissue microstructure.
    • The experiments demonstrate potentially improved contrast and characterization of microstructural changes in gliomas, enhancing the assessment of glioma progression and treatment response in clinical settings.
    • The paper presents good validation of the proposed RDSI method using both ex vivo monkey dMRI data and in vivo human data, supporting the generalizability and applicability of the technique.
  • 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 method relies on a relatively complex mathematical model, which could make it more challenging for non-experts to understand and implement.
    • The paper does not provide a detailed comparison of RDSI with other state-of-the-art techniques in terms of quantifying the accuracy and reproducibility of the derived microstructural parameters, limiting the assessment of the method’s performance.
    • The in vivo human data used in the study includes only one healthy subject and three patients with gliomas, which may not be enough to comprehensively evaluate the method’s robustness and generalizability across different populations and pathological conditions.
  • 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

    The code will be shared if accepted.

  • Please provide detailed and constructive comments for the authors. Please also refer to our Reviewer’s guide on what makes a good review: https://conferences.miccai.org/2023/en/REVIEWER-GUIDELINES.html
    • The authors may consider expanding the evaluation to include other methods besides REDIM
    • I wonder how is the diffusion image acquisition protocol is related to the precision of the estimated parameters. To what extent does it degrade as the protocol is simplified?
    • There is not a comparison made with the glioma data, so I cannot tell if the glioma results something beyond simpler diffusion modeling approaches (or even simply the T1)
    • The figure captions could provide more detail.
    • Typo in abstract “relaxion”
    • The authors may consider citing and discussing this following paper:

    Veraart, Jelle, Dmitry S. Novikov, and Els Fieremans. “TE dependent Diffusion Imaging (TEdDI) distinguishes between compartmental T2 relaxation times.” NeuroImage 182 (2018): 360-369.

  • Rate the paper on a scale of 1-8, 8 being the strongest (8-5: accept; 4-1: reject). Spreading the score helps create a distribution for decision-making

    5

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

    The paper presents a useful and rigorously developed approach, although the evaluation is somewhat lacking, so I am suggesting a weak accept.

  • Reviewer confidence

    Confident but not absolutely certain

  • [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

    This paper presents a relaxation-diffusion model applied to diffusion data with different TEs that characterizes tissue apparent relaxation coefficient for a spectrum of diffusion length scales, factors out intra-voxel orientation heterogeneity (RDSI) and shows results in in-vivo healthy data, glioblastoma data and monkey ex vivo 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.

    Main strengths: -certainly an interesting and actively research area (multi-parametric, multi contrast MRI) -lovely data available, including quite valuabel glioblastom and monkey data with multiple TEs

  • 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.

    Main weaknesses: -evaluation! There are lots of claims made re what the models shows with very little evaluation. The authors have access to ex-vivo data and potentially for the glioblastoma patients even biopsy tissue (?) and must be able to come up with better ways to put their findings into context and properly justify the claims! -details- lots of important details eg on the acquisition are missing, this makes it really hard to understand the methods and the contributions here.

  • Please rate the clarity and organization of this paper

    Satisfactory

  • 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

    They promise to make data and code 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://conferences.miccai.org/2023/en/REVIEWER-GUIDELINES.html

    Abstract: the abstract does not actually describe well what the content of the paper is.It says it was examined in “an in vivo dataset, [..] involving different health conditions” - however the paper is quite glioblastoma focused and has in addition ex-vivo monkey data?

    Methods: more details on the acquisition are required & 21 min sounds almost too fast! (eg multiple-spin-echos after one weighting? or separate acquisitions for each TE?, nr of slices, full brain coverage? It reads like it was acquire in separate acquisitions during the description of the processing (“To compensate for motion, data with different TEs were preprocessed separately and then aligned using rigid registration based on the images without diffusion weighting.”)

    • Also, the acquisitions with different TE (assuming for now this is how they are acquired), how do they vary in diffusion time or was a sequence implemented to assure the deltas remain constant? This is very relevant given the explicite inclusion of diffusion length scales etc..

    Results: -Beautiful images in Fig 1 and 2! Zooms would be nice, eg into the gliomas in (b). Can the scales be bigger? Really hard to appreciate right now.

    -Figure 4d - the areas around the gliomas mainly look as if they hit the upper bound (Cov NR >=0.3). Is this a display or indeed an analysis issue?

    -Section 3.1: “REDIM overestimates the anisotropic volume fraction..” -not sure this claim is actually proven here! Maybe RDSI underestimates it? Given that this is monkey ex-vivo tissue, is there any gold standard (eg by histology, ..) available for these values? -Same here “Our results indicate that MTE dMRI is more sensitive to neurite morphology than STE dMRI” - Given the lack of gold standard / ground truth, hard to make this claim! -And finally here “RDSI improves the detection of small metastases, delineation of tumor extent, characterization of the intratumoral microenvironment when compared to conventional microstructure models (Fig. 4(c)).” - this is a too strong claim given the evidence presented here and the lack of ground truth and quantitative or radiological evaluation.

    -Also the derivations made, eg on necrosis and microvascularity (“RDSI provides useful information on microvascularity and necrosis helpful for facilitating early stratification of patients with gliomas (Fig. 4(d)).”) are too strong and not sufficiently validated/explained/put into context.

    -“It is apparent that at higher b-values, A greater fractions of voxels in the restricted compartment have relaxation times within 100 to 200 ms, particularly for higher-grade gliomas.” - I am just not sure what this is actually telling us? Why this focus on the range 100-200ms?

    Conclusion/Discussion: -There is no discussion and no limitation section etc. -I am really lacking context for the findings, quantitative evaluation and more thorough thought processes outlining how the claims were reached.

  • Rate the paper on a scale of 1-8, 8 being the strongest (8-5: accept; 4-1: reject). Spreading the score helps create a distribution for decision-making

    4

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

    I think this could be quite a nice paper but is not at the moment in a state (in my opinion) where it would be very valuable to the readers and MICCAI attendees. Details are missing, evaluation is missing, limitations are missing etc and I somehow doubt this can all be addressed in rebuttal.

  • Reviewer confidence

    Confident but not absolutely certain

  • [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 proposed a model called “relaxation-diffusion spectrum imaging (RDSI)” to estimate T2 relaxation times and diffusion parameters within different brain tissue compartments, based on multiple echo-time (MTE) diffusion imaging data. In the RDSI model, multi-compartment model was simplified by applying the spherical mean technique which factors out fiber orientation information. The authors validated the RDSI model using ex vivo monkey and in vivo clinical data, and the results showed that RDSI-derived measures on relaxation and diffusion were compartment-specific, suggesting diagnostic potential in capturing tissue abnormalities.

  • 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.
    • Novel formulation for characterising both the T2 relaxation and diffusion properties for different brain tissues.
    • Strong evaluation using both ex vivo monkey and in vivo clinical data.
    • Clear demonstration of the clinical feasibility and the diagnostic potential of the model.
  • 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.
    • Not very clear on data processing and analyses aspects, e.g., how measures of relaxation rates and diffusivity were extracted from different brain tissues (white matter, subcortical gray matter, and cortical gray matter), and how those data were analysed.
  • 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

    It would be good to disclose the version of the OSQP solver and what language interface was used. All the software and tools involved for the diffusion data processing should be clearly described.

  • Please provide detailed and constructive comments for the authors. Please also refer to our Reviewer’s guide on what makes a good review: https://conferences.miccai.org/2023/en/REVIEWER-GUIDELINES.html
    • Abstract: fix the typo “relaxion-diffusion”
    • Section 2.4: The “std()” in Mean NR’s equation is confusing, is it mean or standard deviation?
    • Section 3.2: In the 1st paragraph, the last sentence refers to Fig. 2b (not Fig. 1b). In the 2nd paragraph, fix the capitalized A.
    • Section 3.5: It is not very clear why fiber bundles with slower relaxation times contribute less to diffusion signals acquired with a longer TE. This sentence is a bit convoluted. Better to explain in the context of white matter, subcortical gray matter, and cortical gray matter. In addition, it would make this paper more interesting to show reconstructed fODFs for the glioma subjects.
    • Fig. 2. explain what the diagram in the middle means in caption. “Restriced” is hard to read in the grayscale color bar, consider switching to white color.
    • Fig. 3. Clarify clarify what the arrows stand for in caption.
    • Fig. 4b. Similar to Fig. 2, consider using white text for “Long TE” to improve readability.
    • Fig. 6 To improve accessibility of color-blind readers, indicate for each row which superficial WM regions it refers to. Explain what the arrows are pointing at in caption.
  • Rate the paper on a scale of 1-8, 8 being the strongest (8-5: accept; 4-1: reject). Spreading the score helps create a distribution for decision-making

    6

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

    This is a good paper with moderate weakness. The authors addressed the problem of accurate characterisation of tissue microstructure, an interesting and important topic in translating advanced diffusion MRI techniques to the clinics. The simplification of the more complex multi-compartment model was reasonably justified. The methods of estimating both the relaxation and diffusion parameters were clearly described. Evaluations were performed in both ex vivo monkey and in vivo clinical data, and the results clearly suggested that incorporating relaxation produced distinguishable profiles for different brain tissue compartments. Nevertheless, the overall clarity of this paper need to be improved; providing details for data processing and analyses will help validation and reproducibility, bringing the approach closer to clinical adoption. The readability and accessibility for some of the figures can also be improved (see detailed comments).

  • Reviewer confidence

    Confident but not absolutely certain

  • [Post rebuttal] After reading the author’s rebuttal, state your overall opinion of the paper if it has been changed

    6

  • [Post rebuttal] Please justify your decision

    Detailed responses were provided to address the main points raised by the reviewers and meta-reviewer.



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 paper focuses on an interesting and important topic that may benefit the translation of advanced diffusion MRI to the clinics, and it has proposed a novel formulation for disentangling T2 relaxation and diffusion properties. All reviewers agree that the validation is impressive with both ex vivo monkey data and in vivo human data. However, there are also some weaknesses that should be addressed. For example, data acquisition and processing details are missing, some claims should be properly justified, and the limited number of subjects needs to be discussed.



Author Feedback

#Data protocol and processing (MetaRev, R1, R2, and R3). The seven dMRIs with different TEs were separately obtained using a spin-echo echo-planar imaging sequence (voxel size = 1.5mm^3, image size = 160 x 160, 96 slices, whole-brain coverage, acceleration factor = 3) with fixed TR and diffusion time. All dMRIs were first preprocessed separately and then aligned using rigid registration based on the non-diffusion-weighted images (Sec. 2.5). The lowest TE was set to minimize the contribution of the myelin water to the measured signal and the largest TE was chosen as a trade-off between image contrast and noise. Following previous studies [5,14] and in-house testing, we used a spectrum of TE scales from 75 to 135 ms to cover tissue heterogeneity. We used MRtrix to generate tissue segmentations (cortical and subcortical GM, WM, CSF, and pathological tissue) based on the T1w data. To avoid tissue-type misclassifications due to WM lesions, we used 5ttedit to add the previously determined lesion maps to the 5TT files using the optional 5th tissue type (pathological tissue).

#Complexity of Mathematical Models (R1). Unlike existing techniques that require computationally expensive non-linear procedures to fit models to the data, we formulated the rdMRI fitting framework as a linear system (Eqs. 4 and 6) that can be conveniently solved using fast algorithms. We demonstrated this linearization of the fitting problem by providing an alternate parameter estimation method; moreover, the RDSI framework is general and flexible enough to work for a wide range of microstructure imaging problems, e.g., [6, 24].

#Limitation of sample size (MetaRev, R1). We proposed a novel framework to analyze rdMRI for probing tissue microstructure in heterogeneous environments. For this, we evaluated RDSI using both ex vivo monkey and in vivo human brain MTE data. With an in vivo dataset, we showed the diagnostic potential of RDSI in differentiating tumors and normal tissues (Figs. 2-4). Since rdMRI datasets are not publicly available, detailed evaluation is naturally limited in this work.

#Comparison (R1 and R2). We compared our model with the publicly available model REDIM to estimate T2-independent parameters and found that our model is competitive (Fig.1). We cannot compare our model with other existing methods as their codes are unavailable. Moreover, previous studies did not consider signal modeling by characterizing the apparent relaxation rate related to dMRI with different b-values, components, and health conditions.

#Some claims should be properly justified.

  1. Ex vivo. In-line with the results from [17], Fig.1(a) suggests that REDIM overestimates the anisotropic volume fraction compared with RDSI, and REDIM uFA shows blurred boundaries between the GM and superficial WM.

  2. In vivo. Our results suggest that at higher b-values many voxels in the restricted compartment have relaxation times within 100 to 200 ms, particularly for higher-grade gliomas. There has been evidence that this might be related to prolonged transverse relaxation time due to increased water content within the tumor (Upadhyay and Waldman, 2011; Ellingson et al., 2017; Hu et al., 2020; Li J. et al., 2020). This property is useful in the visualization of peritumoral edema, an area containing infiltrating tumor cells and increased extracellular water due to plasma fluid leakage from aberrant tumor capillaries that surrounds the tumor core in higher-grade gliomas.

  3. In vivo. Fig.4(a) shows the relaxation times of the restricted compartment in WM lesions, indicating that relaxation times are longer in gliomas than normal WM tissue. The higher T2 in grade 4 glioma is associated with changes in metabolite compositions, resulting in remarkable changes in neurite morphology in lesioned tissues (Fig.4(c-d)), consistent with previous observations [10, 21]. The rate of longitudinal relaxation time has been shown to be positively correlated with myelin content.



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 authors have properly addressed the questions of the reviewers. In particular, the data acquisition and processing details are clarified, and the claims are better justified. I believe this is an excellent work that can be accepted.



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 thought that the proposed Relaxation-Diffusion Spectrum Imaging (RDSI) framework is of interest to the MICCAI community, sufficiently novel, and sufficiently well evaluated for presentation at the main conference. There were some concerns regarding missing details, especially with respect to acquisition parameters, and a missing discussion of remaining limitations. Based on the rebuttal, I am confident that these will be added to the final version. I agree with the reviewers that a more comprehensive comparison with respect to existing methods and findings from histology would further strengthen this work, but believe that this might exceed the scope of a conference paper, and might rather be kept in mind for a potential extension towards a full journal paper.



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.

    Joint relaxation-diffusion MRI has been increasingly recognized as an important task in the MR field. The authors have proposed a novel method for doing so. Even thought there are several minor concerns from the reviewers, the overall contribution of the paper is high. The additional AC suggests acceptance of the paper at MICCAI.



back to top