Authors

Khoi Minh Huynh, Ye Wu, Sahar Ahmad, Pew-Thian Yap

Abstract

Most diffusion biophysical models capture basic properties of tissue microstructure, such as diffusivity and anisotropy. More realistic models that reate the diffusion-weighted signal to cell size and membrane permeability often require simplifying assumptions such as short gradient pulse and Gaussian phase distribution, leading to statistical features relating to tissue properties instead of real quantitative measurements. Here, we propose a method to quantify tissue microstructure without jeopardizing accuracy owing to unrealistic assumptions. Our method utilize realistic signals simulated from the geometries of cellular microenvironments as fingerprints, which are then employed in a spherical mean estimation framework to disentangle the effects of orientation dispersion from the microscopic tissue properties. We demonstrate the efficacy of microstructure fingerprinting in estimating intra-cellular, extra-cellular, and intra-soma volume fractions as well as axon radius, soma radius, and membrane permeability.

Link to paper

DOI: https://doi.org/10.1007/978-3-031-43993-3_13

SharedIt: https://rdcu.be/dnwNe

Link to the code repository

N/A

Link to the dataset(s)

N/A

Reviews

Review #2

Please describe the contribution of the paper

This paper presents a microstructure fingerprinting technique that aims to accurately measure tissue properties such as axonal and somatic cell size and permeability beyond traditional diffusivity and anisotropy measurements, with results showing agreement with ex-vivo histological measurements.
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.
- Presents a novel microstructure fingerprinting technique (MF-SMSI) that can estimate tissue properties beyond diffusivity and anisotropy, allowing quantification of cell size and permeability associated with axons and somas.
- Shows the effectiveness of the MF-SMSI technique by comparing its results with histological samples and previous studies, and it provides detailed validation through in-silico and in-vivo experiments.
- Highlights the biases associated with previous methods and how MF-SMSI addresses those biases, making it a more realistic measurement of axonal radii.
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.
- Validation experiments were primarily based on synthetic data, which may not fully represent the complexity of in vivo biological tissue.
- The model used assumes simplified geometries for the soma and axon compartments, which may not capture the full range of anatomical variations.
- The method requires a large dictionary of fingerprints to be generated beforehand, which may limit its generalizability to different acquisition protocols or populations, and possibly interactions among them.
- The paper does not compare different types of brain tissues, e.g. WM vs. cerebellum vs. cortex vs subcortical, which have well known differences in tissue microstructure.
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

Most of the tools used are publicly available, and the authors indicate they will share their specific code.
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
- It would be helpful to show axon diameter estimates in gray matter. Axons do exist there as well.
- This work seems to assume that all cells have same intra-cellular diffusivity, but is this the case?
- In Fig 4, why is the spatial extent of the large soma radius so much larger than soma volume fraction?
- Intra-voxel heterogeneity of microstructure parameters is not accounted for by the model, and this should be mentioned as a limitation.
- In Fig 4, the caption doesn’t match the image. v_EC is listed twice (one should be v_CSF), and there are no top and bottom panels.
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?

I recommend a weak accept because the paper is interesting and well-executed; however, there are some limitations of the approach and the evaluation experiments.
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 #5

Please describe the contribution of the paper

Authors present a microstructure fingerprinting method that can provide accurate and reliable measurements of tissue properties. The proposed method allows quantification of cell size and permeability associated with axons and somas.
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. Authors propose a novel method to quantify tissue microstructure without unrealistic assumptions.
2. Authors demonstrate the efficacy of their proposed microstructure fingerprinting in estimating intra-cellular, extra-cellular, and intra-soma volume fractions as well as axon radius, soma radius, and membrane permeability.
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 size and type of the in-vivo test dataset are limited.
2. Only HCP-like parameters are used in testing.
Please rate the clarity and organization of this paper

Excellent
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 used dataset is public, but not sure if the authors will release their code.
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
1. Test on more datasests with different populations (e.g., babies, children, older adults, patients, etc)
2. Test on more acquisition parameter settings, instead of only the HCP setting.
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?

The proposed microstructure fingerprinting is interesting, but more datasets and settings are needed for testing its effectiveness.
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 #4

Please describe the contribution of the paper

The paper presents a microstructure fingerprinting estimation method that enables quantifying different parameters of the environment, such as the axon radius or soma radius. The Authors simulate the local environment using a biologically-inspired in silico data and use in vivo Human Connectome Project 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.

The paper has a strong motivation and is appropriately written. The organization of the paper follows the requirement of the MICCAI main track. The paper has a strong evaluation with with in vivo data. The experimental results are precise, and the figures are generated in vector-like graphics.
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 doubts regarding the actual contribution of the paper as the Authors primarily integrate already known tools and methods from the literature.
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 reproducibility of the paper is high as the Authors provide in-depth details of the methods (experimental setups, etc..).
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

Please, correct the unit in “For b ≤ 3000ms”. Please, define the SNR of the signal.
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?

Overall, the paper is above the MICCAI main track threshold. It is clearly written, with all important references included therein. The Authors carefully refer to the literature once simulating the in silico data.
Reviewer confidence

Somewhat confident
[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

The paper proposed a new method to quantify tissue microstructure without jeopardizing accuracy owing to unrealistic assumptions and demonstrated the efficacy of microstructure fingerprinting.
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 useful and novel, and it has been evaluated with solid experiments. The paper is well written. The figures look good and informative.
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.

It does not provide a ccomprehensive literature review in the introduction part.
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

Reproducibility is relatively low for this paper. THe code is not provided.
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

Give a more detailed description about the current methods for microstructure fingerprinting.
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 method is novel and solidly evaluated.
Reviewer confidence

Not confident
[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

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.

All reviewers consent that the proposed method is novel and its efficacy is demonstrated. The reviewers also suggest several possibilities of extending the experiments for validation, for example, with more datasets beyond synthetic data. Besides, some methodological limitations are pointed out and should be discussed.

Author Feedback

We thank the meta reviewer and reviewers for the helpful comments, which we address as follows:

-Dictionary construction “The method requires a large dictionary of fingerprints to be generated beforehand, which may limit its generalizability to different acquisition protocols or populations, and possibly interactions among them.” By solving the BT-PDE instead of using Monte-Carlo simulation, our method only requires the dictionary to be constructed once for each acquisition protocol, taking only 5 minutes. Thus, our method is generalizable to different protocols.

-Diffusivity assumption “This work seems to assume that all cells have same intra-cellular diffusivity, but is this the case?” For the axon and soma (intra-cellular) compartment, we use a realistic configuration defined by a wide range of radius and permeability to simulate the signal, resulting in varying apparent diffusivity. The simulation assumes that, without obstacles or membranes, water molecules diffuse at same speed in the brain as dMRI signal is sensitive to obstacles rather than to cell cytoplasm (Rorschach HE, Lin C, Hazlewood CF. Diffusion of water in biological tissues. Scanning Microscopy. 1991;1991(5):1.) . For extra-cellular and free-water compartments, we use a spectrum of diffusivity covering biological relevant values in the human brain.

-Volume fraction and radius map “In Fig 4, why is the spatial extent of the large soma radius so much larger than soma volume fraction?” In some voxels (notably in the white matter), the soma volume fraction is small (not visible but non-zero), causing the volume fraction map to appear smaller than the radius map.

-Intra-voxel heterogeneity “Intra-voxel heterogeneity of microstructure parameters is not accounted for by the model, and this should be mentioned as a limitation.” By using the spherical mean, we remove the confounding effect of fiber dispersion. And by using a diffusion spectrum accounting for differences in radius, permeability, extra-cellular matrix, and free-water diffusion, our model accounts for intra-voxel heterogeneity.

-Testing on more datasets “Test on more datasets with different populations (e.g., babies, children, older adults, patients, etc)” “Test on more acquisition parameter settings, instead of only the HCP setting.” “The size and type of the in-vivo test dataset are limited.” “Only HCP-like parameters are used in testing.” We used data of 35 adult subjects from the HCP to validate our method as there are multiple in-vivo studies using this cohort and histological results from subjects in a similar age range, making comparison and validation straightforward. We have tested the generalizability of our method to various protocols or datasets covering the entire human lifespan.

back to top

Microstructure Fingerprinting for Heterogeneously Oriented Tissue Microenvironments