List of Papers Browse by Subject Areas Author List
Abstract
Accurate 3D+t whole-heart mesh reconstruction from cine MRI is a clinically crucial yet technically challenging task. The difficulty of this task arises from two coupled factors: inherently sparse sampling of 3D cardiac anatomy by 2D image slices and the tight coupling between cardiac shape and motion. Current cardiac image-to-mesh approaches typically reconstruct only a subset of cardiac chambers or a single phase of the cardiac cycle. In this work, we propose CineMesh4D, a novel end-to-end 4D (3D+t) pipeline that directly reconstructs patient-specific whole-heart mesh from multi-view 2D cine MRI via cross-domain mapping. Specifically, we introduce a differentiable rendering loss that enables supervision of 3D+t whole-heart mesh from multi-view sparse contours of cine MRI. Furthermore, we develop a dual-context temporal block that fuses global and local cardiac temporal information to capture high-dimensional sequential patterns. In quantitative and qualitative evaluations, CineMesh4D outperforms existing approaches in terms of reconstruction quality and motion consistency, providing a practical pathway for personalized real-time cardiac assessment. The code will be publicly released once the manuscript is accepted.
Links to Paper and Supplementary Materials
Main Paper (Open Access Version): https://papers.miccai.org/miccai-2026/paper/0099_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{LiuXia_CineMesh4D_MICCAI2026,
author = { Liu, Xiaoyue AND Yuan, Xiaohan AND Chan, Mark Y. AND Sia, Ching-Hui AND Li, Lei},
title = { { CineMesh4D: Personalized 4D Whole Heart Reconstruction from Sparse Cine MRI } },
booktitle = {Medical Image Computing and Computer Assisted Intervention -- MICCAI 2026},
year = {2026},
publisher = {Springer Nature Switzerland},
volume = {LNCS 16889},
month = {September},
page = {pending}
}
Reviews
Review #1
- Please describe the contribution of the paper
This paper proposes a framework for reconstructing the 4D whole-heart mesh from sparse multi-view cine MRI. The main contribution is not just the final network, but the overall framework which combines direct cine-to-mesh mapping, a contour-based differentiable rendering loss to supervise the reconstruction from sparse 2D views, and a temporal module to improve motion consistency across the cardiac cycle. In my view, the paper’s main contribution is a practical 4D whole-heart reconstruction pipeline from routine cine MRI.
- 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 main strengths of the paper are that it addresses an important problem, namely 4D whole-heart reconstruction from sparse multi-view cine MRI, and that it does so with a fairly complete end-to-end pipeline rather than focusing only on a single chamber or a single cardiac phase. The paper also explicitly models temporal coherence, which is appropriate for a 4D task, and the contour-based differentiable rendering supervision is a reasonable way to use sparse 2D cine views. Finally, the evaluation is reasonably broad, including geometric, contour, and temporal metrics as well as ablations.
- 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.
Weaknesses also include the broad but insufficiently demonstrated clinical value, the under-described reference mesh generation process, and the relatively weak qualitative evaluation. The paper emphasizes “personalized” and “functional” reconstruction, but does not clearly define the clinical workflow or evaluate functional endpoints such as ejection fraction or EDV/ESV. In addition, the reference mesh generation pipeline is only described at a high level, which limits reproducibility and makes the reported errors harder to interpret. Finally, Fig. 3 and the supplementary video mainly show reconstructed meshes without clear side-by-side comparison to the reference mesh, making qualitative assessment less convincing.
- Please rate the clarity and organization of this paper
Satisfactory
- 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 authors claimed to release the source code and/or dataset upon acceptance of the submission.
- Based on your review and your understanding of the MICCAI Scientific Code of Ethics, do you believe this submission may involve a potential ethics concern or violation?
N/A
- 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/2026/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 — marginally above the acceptance threshold, but would not mind if rejected, dependent on rebuttal
- Please justify your recommendation. What were the major factors that led you to your overall score for this paper?
This paper addresses an important problem and presents a reasonably complete 4D whole-heart reconstruction system from multi-view cine MRI. The paper goes beyond partial-heart or single-phase reconstruction, and the overall framework is technically coherent, with temporal modeling, contour-based supervision, and fairly broad quantitative evaluation. These aspects make it a useful and relevant application-oriented contribution.
At the same time, I do not view it as a strong methods paper. The methodological novelty is moderate, the clinical value is discussed more broadly than it is demonstrated, and some parts of the evaluation could be stronger, especially with respect to functional metrics, qualitative comparison to reference meshes, and the description of how the reference meshes are generated. These concerns lower my enthusiasm.
Overall, however, I found the work technically solid, practically motivated, and sufficiently mature to be of interest to the MICCAI community. My score reflects that I see it as slightly above the acceptance threshold as an application/system paper, even though it does not reach the level of a highly novel methodological contribution.
- 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
The paper proposes CineMesh4D, an end to end system that reconstructs a 4D whole heart mesh (3D + time) directly from sparse multi view cine MRI. The method avoids intermediate segmentation and volumetric reconstruction by learning a direct image→mesh mapping. They designed a novel differentiable rendering loss inspired by a Beer–Lambert-Law and a dual-context temporal block for global/local motion modeling.
- 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) Differentiated technical contribution: Beer–Lambert differentiable rendering loss is a principled, novel and physically motivated approach to supervise 3D mesh reconstruction from 2D contours.
2) The dual-context temporal block is good contribution and it is properly ablated. The design explicitly fuses global motion descriptors with local sliding-window temporal context, solving a common problem of jittery framewise predictions which proved by results showing lower mesh jitter and cleaner volume curves.
3) First end-to-end 4D whole-heart approach from multi-view sparse cine MRI without intermediate volumetric segmentation is actually a genuine gap in the literature. Most prior mesh reconstruction works focus on LV or biventricular models. This paper reconstructs LV, RV, LA, RA, myocardium, improving clinical relevance.
4) Good integration of classical and modern ideas: MeshVAE + GCN + Exphormer + LoRA + differentiable rendering + temporal modeling all support the central goal effectively without overengineering. The system feels balanced and practical.
5) Ablation across view combinations, pretraining strategy, and temporal block is comprehensive.
6) The rendering loss and dual context temporal block are a novel, well designed and goal focused contributions.
- 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) No comparison on a publicly available benchmark, all baselines are re-implemented without cross-study reproducibility, would like to see the blind testing on publicly available datasets like MMWHS, by using multi-planar reformation (mrp) the long and short axis views can be generated by this dataset which has whole heart segmentation anyway so it could prove generalizability of the proposed method.
2) The proposed method requires atlas based dense 3D segmentation to train MeshVAE. Training ground truth meshes come from a multi-step pipeline: 2D segmentations → atlas deformation → 3D segmentation → template mesh warping. This adds noise and biases that might limit generalization, yet little discussion is provided.
3) The paper notes that atrial reconstruction is less accurate due to sparse sampling. A more explicit failure-case analysis highlighting where the proposed method succeeds vs rest will show how much this method impacts and would strengthen the paper.
4) Since the loss compares against 2D ground-truth contours, any segmentation inconsistencies or label noise may misguide training. The paper doesn’t quantify sensitivity to segmentation errors.
- 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 authors claimed to release the source code and/or dataset upon acceptance of the submission.
- Based on your review and your understanding of the MICCAI Scientific Code of Ethics, do you believe this submission may involve a potential ethics concern or violation?
N/A
- 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/2026/en/REVIEWER-GUIDELINES.html
For questions refer to weaknesses
- 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 — marginally above the acceptance threshold, but would not mind if rejected, dependent on rebuttal
- Please justify your recommendation. What were the major factors that led you to your overall score for this paper?
Technically novel with a principled rendering loss and strong temporal modeling. The main limitation is dataset scale and lack of public benchmark validation, which should be addressed.
- 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
Review #3
- Please describe the contribution of the paper
This paper proposes an end-to-end pipeline for 3D+t whole heart mesh reconstruction from multi-view 2D cine MRI. Specifically, a differentiable rendering loss is introduced to enable 3D mesh supervision from 2D contours. A dual-context temporal block to fuse global and local temporal features. The proposed pipeline was trained end-to-end, allowing the shape reconstruction and temporal consistency to be jointly optimised.
- 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) The task this paper aims for is clinically meaningful for cardiac digital twins. (2) The results are strong, particularly on the temporal smoothness improvements.
- 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) The size of the dataset is a bit small, with only 57 test cases. (2) Insufficient literature review and comparison (3) The technique novelty should be further justified.
- 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 authors claimed to release the source code and/or dataset upon acceptance of the submission.
- Based on your review and your understanding of the MICCAI Scientific Code of Ethics, do you believe this submission may involve a potential ethics concern or violation?
N/A
- 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/2026/en/REVIEWER-GUIDELINES.html
(1) Many recent papers fail to cite or compare against, of which also works on 4D whole heart reconstruction from multi-view cine-MR. The technical novelty of this paper should be further justified.
Liu, et al. “Personalized 4D Whole Heart Geometry Reconstruction from Cine MRI for Cardiac Digital Twins.” International Workshop on Digital Twin for Healthcare, 2025.Qiao, et al. “Mesh4D: A Motion-Aware Multi-view Variational Autoencoder for 3D+ t Mesh Reconstruction.” MICCAI, 2025.Xia, et al. “Automatic 3D+ t four-chamber CMR quantification of the UK biobank: integrating imaging and non-imaging data priors at scale.” Medical Image Analysis 80 (2022): 102498. (2) In DeepMesh (ref. 13 in the draft), it also proposed a differentiable mesh-to-image rasterizer to enable training the network using a binary mask. What is the benefit of the proposed Beer–Lambert-based rendering loss?
(3) The primary compared alternative is HybridVNet, which is not designed for the full cardiac cycle. The paper would benefit from a comparison against the aforementioned prior methods.
- 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.
(3) Weak Reject — marginally below the acceptance threshold, but would not mind if accepted, dependent on rebuttal
- Please justify your recommendation. What were the major factors that led you to your overall score for this paper?
Overall, this paper is good. However, many related papers are not discussed or compared, making it hard to justify the paper’s technical novelty.
- 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
- Please justify your decision. 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 reviewers generally view this as a solid and relevant contribution, with consistent strengths in problem formulation, technical design, and overall system completeness. In particular, the end-to-end 4D whole-heart reconstruction pipeline and the integration of temporal modeling and differentiable rendering are seen as meaningful steps toward practical clinical use. At the same time, several concerns are raised across reviews. These mainly relate to limited validation on public benchmarks, unclear generalization due to reliance on a multi-stage reference mesh pipeline, and gaps in positioning against recent prior work. Some reviewers also question whether the level of methodological novelty fully matches the claims, especially given similarities to existing approaches. I generally agree with the reviewers’ assessment. The paper is technically sound but the concerns around evaluation, benchmarking, and positioning are important and should be addressed by the authors to strengthen the submission.
