List of Papers Browse by Subject Areas Author List
Abstract
Function-structure connectivity (FSC) coupling helps reveal alterations in the interplay between brain functional connectivity (FC) and structural connectivity (SC) caused by neurocognitive decline. Existing studies on FSC coupling typically focus on modeling interactions between static FC and SC features, ignoring temporal dynamics conveyed in functional MRI (fMRI) time series. Additionally, conventional strategies often compute global whole-brain FSC correlation or assess local region-specific FSC correspondences, without capturing complex inter-region dependencies between FC and SC patterns. To this end, we propose a dynamic function-structure connectivity coupling (DFSC) framework to predict progression trajectories in neurocognitive decline with fMRI and diffusion tensor imaging (DTI) data. In DFSC, we first construct static SC and dynamic FC graphs and use graph neural networks (GNNs) for feature learning, yielding new SC and FC embeddings. Based on these embeddings, we construct dynamic local-to-global FSC coupling graphs to capture both region-specific and inter-region dependencies between FC and SC, followed by GNNs to generate dynamic FSC coupling embeddings. These multi-view embeddings are finally fed into a squeeze-excitation readout module and a Transformer for feature fusion and prediction. Experimental results on two datasets with paired fMRI and DTI data from a total of 231 subjects demonstrate that our DFSC outperforms several state-of-the-art methods. With the DFSC, one can identify both discriminative brain regions and between-group FSC coupling difference, facilitating objective quantification of structural and functional brain changes associated with neurocognitive decline.
Links to Paper and Supplementary Materials
Main Paper (Open Access Version): https://papers.miccai.org/miccai-2025/paper/3226_paper.pdf
SharedIt Link: Not yet available
SpringerLink (DOI): Not yet available
Supplementary Material: Not Submitted
Link to the Code Repository
N/A
Link to the Dataset(s)
N/A
BibTex
@InProceedings{WanQia_Dynamic_MICCAI2025,
author = { Wang, Qianqian and Wang, Wei and Li, Hong-Jun and Lin, Weili and Liu, Mingxia},
title = { { Dynamic Function-Structure Connectivity Coupling for Predicting Progression Trajectories in Neurocognitive Decline } },
booktitle = {proceedings of Medical Image Computing and Computer Assisted Intervention -- MICCAI 2025},
year = {2025},
publisher = {Springer Nature Switzerland},
volume = {LNCS 15971},
month = {September},
page = {299 -- 309}
}
Reviews
Review #1
- Please describe the contribution of the paper
The paper proposes a neural network architecture called Dynamic Function-Structure Connectivity Coupling (DFSC), designed to model both region-specific and inter-regional dependencies between dynamic functional connectivity (FC) and structural connectivity (SC). By training the network to predict disease labels using paired fMRI and DTI data from the ADNI and HCD datasets, DFSC achieves strong classification performance. Moreover, it identifies disease-relevant regions of interest (ROIs) and reveals alterations in function-structure coupling (FSC) associated with neurocognitive decline.
- Please list the major strengths of the paper: you should highlight 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 addresses an underexplored aspect of FC-SC coupling—region-specific coupling between dynamic FC and SC.
- The paper proposes a novel neural network architecture, DFSC, to model region-specific dynamic FC-SC coupling. The design incorporates an explicit FSC coupling operation (using Pearson correlation between the learned FC and SC features) and a squeeze-excitation mechanism to transform node-level embeddings into graph-level representations. This enables the model to identify specific coupling patterns and ROIs that play key roles in disease prediction.
- In addition to comprehensive baseline comparisons and ablation studies, the authors also highlight specific FSC coupling patterns and ROIs that are discriminative between subject groups.
- Please list the major weaknesses of the paper. Please provide details: for instance, if you state that a formulation, way of using data, demonstration of clinical feasibility, or application is not novel, then you must provide specific references to prior work.
- While the authors claim that the proposed method captures “local-to-global” FSC coupling graphs, this concept is not clearly reflected in the method section. The architecture primarily computes Pearson correlations between node-level FC and SC embeddings, which suggests that the modeled coupling is region-specific, or local-to-local. There is no explicit mechanism or architectural component that aggregates or reasons over global-level interactions, making the “local-to-global” claim appear unsubstantiated.
- In the Multi-View Feature Fusion & Prediction module, features from different time points and modalities (FC, SC, and FSC) are integrated using a self-attention mechanism. However, it is unclear whether positional encoding was used to distinguish between features from different time segments or modalities.
- The paper reports the top 10 discriminative ROIs and between-group FSC coupling differences. However, the robustness of these findings is unclear. It remains to be seen whether variations in neural network initialization or different subject groupings (e.g., from different cross-validation folds) would substantially alter the identified ROIs or coupling patterns. Providing measures of variability or consistency across runs would strengthen the reliability of these interpretations.
- 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.
- Optional: If you have any additional comments to share with the authors, please provide them here. Please also refer to our Reviewer’s guide on what makes a good review and pay specific attention to the different assessment criteria for the different paper categories: https://conferences.miccai.org/2025/en/REVIEWER-GUIDELINES.html
N/A
- 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.
(4) Weak Accept — could be accepted, dependent on rebuttal
- Please justify your recommendation. What were the major factors that led you to your overall score for this paper?
The paper addresses a novel and underexplored problem—region-specific dynamic FC-SC coupling—with a thoughtfully designed neural architecture. The method shows strong empirical performance and includes solid baseline comparisons and ablation studies. However, some claims (e.g., “local-to-global” coupling) lack clear support, and the robustness of the interpretability findings is not well validated.
- Reviewer confidence
Confident but not absolutely certain (3)
- [Post rebuttal] After reading the authors’ rebuttal, please state your final opinion of the paper.
N/A
- [Post rebuttal] Please justify your final decision from above.
N/A
Review #2
- Please describe the contribution of the paper
Authors propose dynamic function-structure connectivity coupling (DFSC) framework to predict progression trajectories in neurocognitive decline with fMRI and diffusion tensor imaging (DTI). The framework is based on graph neural network and constructs dynamic functional connectivity and static connectivity embeddings, which further are passed to squeeze-excitation readout module and a Transformer for feature fusion and prediction. The approach outperforms several state-of-the-art methods. The main difference from other approaches that that authors use dynamic functional connectivity instead of static one in previous ones.
- Please list the major strengths of the paper: you should highlight 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 proposes a novel multimodal approach that integrates DTI with dynamic functional connectivity, offering a fresh perspective on structure-function interaction. The evaluation is thorough, including comparisons with unimodal baselines and ablation studies. Results are well interpreted, supporting both methodological soundness and clinical relevance.
- Please list the major weaknesses of the paper. Please provide details: for instance, if you state that a formulation, way of using data, demonstration of clinical feasibility, or application is not novel, then you must provide specific references to prior work.
The study uses a relatively small subset of the ADNI dataset — 46 SMC subjects and 48 matched NCs — despite the availability of a substantially larger cohort in ADNI3, including over 300 MCI patients with both fMRI and DTI data. Could the authors clarify the rationale behind using such a limited sample, and whether the proposed method has been tested on the broader multimodal population available within ADNI3?
- 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.
- Optional: If you have any additional comments to share with the authors, please provide them here. Please also refer to our Reviewer’s guide on what makes a good review and pay specific attention to the different assessment criteria for the different paper categories: https://conferences.miccai.org/2025/en/REVIEWER-GUIDELINES.html
N/A
- 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.
(4) Weak Accept — could be accepted, dependent on rebuttal
- Please justify your recommendation. What were the major factors that led you to your overall score for this paper?
Overall assessment: The paper presents an interesting and well-motivated multimodal framework with clear component-wise evaluation and strong benchmarking. Main concern: The primary limitation lies in the use of a small subset of the ADNI dataset (46 SMC and 48 NC subjects), despite the availability of a much larger multimodal cohort in ADNI3. The rationale for this restricted sample selection is not clearly explained. Given the limited size and absence of pretraining, the results are likely affected by high variance and limited generalizability — as also reflected in the large standard deviations reported in the performance tables.
- Reviewer confidence
Confident but not absolutely certain (3)
- [Post rebuttal] After reading the authors’ rebuttal, please state your final opinion of the paper.
N/A
- [Post rebuttal] Please justify your final decision from above.
N/A
Review #3
- Please describe the contribution of the paper
This is an interesting study to utilize dynamic fuctional connectivity (dFC) in combination with structural connectivity (SC) to predict neurocognitive decline. The authors utilize two datasets to test their method and found that the proposed approach is significantly better than the current approach. The investigators have done numerous experiments to show support of their algorithm. Overall, the paper is good.
- Please list the major strengths of the paper: you should highlight 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) Utilizing dFC and dMRI-derived SC and understanding non apriori driven whole brain dFC and SC interactions to study neurocognitive disorders is good. 2) Numerous experiments show the rigor of the research.
- Please list the major weaknesses of the paper. Please provide details: for instance, if you state that a formulation, way of using data, demonstration of clinical feasibility, or application is not novel, then you must provide specific references to prior work.
1) AAL atlas is not a functional atlas and the study might be benefitted by utilizing a functional atlas. Also, the weights of SC is the number of fibers which is sensitive to noise, especially in deterministic tractography. 2) Filtering bottom 70% connection is vague. Do the authors have a rationale for this? 3) The number of features that may come from their algorithm is still significantly greater than the number of participants. How confident are the authors that their algorithm will replicate across different dataset and population? Is it possible to test their algorithm on an independent testing dataset?
- 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 provide sufficient information for reproducibility.
- Optional: If you have any additional comments to share with the authors, please provide them here. Please also refer to our Reviewer’s guide on what makes a good review and pay specific attention to the different assessment criteria for the different paper categories: https://conferences.miccai.org/2025/en/REVIEWER-GUIDELINES.html
N/A
- 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.
(5) Accept — should be accepted, independent of rebuttal
- Please justify your recommendation. What were the major factors that led you to your overall score for this paper?
The paper is innovative in its approach and a good proof-of-concept to develop and validate further.
- Reviewer confidence
Very confident (4)
- [Post rebuttal] After reading the authors’ rebuttal, please state your final opinion of the paper.
N/A
- [Post rebuttal] Please justify your final decision from above.
N/A
Author Feedback
N/A
Meta-Review
Meta-review #1
- Your recommendation
Provisional Accept
- If your recommendation is “Provisional Reject”, then summarize the factors that went into this decision. In case you deviate from the reviewers’ recommendations, explain in detail the reasons why. You do not need to provide a justification for a recommendation of “Provisional Accept” or “Invite for Rebuttal”.
N/A