Pith. sign in

REVIEW 3 major objections 5 minor 42 references

HeartUnloadNet: A Weakly-Supervised Cycle-Consistent Graph Network for Predicting Unloaded Cardiac Geometry from Diastolic States

T0 review · 3 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Pith's one-line read HeartUnloadNet claims the zero-pressure left-ventricular geometry can be predicted straight from the end-diastolic mesh in about 0.02 seconds with sub-millimeter accuracy, outperforming inverse finite-element solvers on 20,700 synthetic…

desk verdict A genuinely novel GNN surrogate for cardiac unloading with thorough ablations, but the headline DSC and 200-sample claims are numerically inconsistent and need correction before the accuracy numbers can be trusted. read the letter →

arxiv 2507.18677 v1 pith:GUFB322Y submitted 2025-07-24 cs.CV physics.med-ph

classification cs.CVphysics.med-ph
keywords cycle-consistentgraphnetworkscardiacinversemodelingweakly-supervisedlearningzero-pressuregeometrybiomechanicalshapepredictionleftventricularmechanicsattentionnetworkfiniteelementsurrogate
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper tries to show that recovering the unloaded, zero-pressure shape of the left ventricle, a step normally done by slow inverse finite-element optimization, can be replaced by a single forward pass of a graph neural network. On synthetic test geometries it reports a node-level overlap score of 0.986 and an average maximum surface deviation of 0.083 cm, with inference around 0.02 seconds per case, about five orders of magnitude faster than the inverse FE baseline and more accurate. The network accepts an end-diastolic mesh of any topology plus physiological parameters such as pressure, stiffness, and fiber orientation, and outputs the displacement to the unloaded state. A cycle-consistency loop lets it train with far fewer labeled examples, keeping overlap near 0.97 with roughly 200 training cases. If correct, this makes a clinically unobservable reference state available in real time for personalized cardiac biomechanics.

What carries the argument

The load-bearing mechanism is a graph-attention encoder coupled with two deformation decoders that share the encoder and run in opposite directions. The encoder builds per-node features from mesh coordinates by aggregating neighbor information with learned attention weights, pools a global mesh representation, and fuses it with an encoded vector of the physiological parameters through cross-attention. One decoder predicts end-diastolic-to-unloaded nodal displacements; the other reconstructs the end-diastolic mesh from the predicted unloaded mesh. The training loss is the supervised squared error on the unloaded mesh plus a weighted cycle term that penalizes mismatch between the reconstructed and original end-diastolic mesh, enforcing reversibility of the deformation and providing weak self-supervision.

What would settle it

Take a heart or high-fidelity model for which the unloaded geometry is known independently, for example an excised ventricle imaged at zero cavity pressure and then loaded to a known pressure, and compare the network's predicted unloaded mesh with that measured state; if the node-level overlap or mean distance is substantially worse than the reported 0.986 and 0.028 cm, the claim that the network predicts true unloaded geometry fails.

Watch

Extended reading notes

Core claim

HeartUnloadNet's central claim is that inferring the unloaded left-ventricular configuration is a learnable graph-to-graph regression: given an end-diastolic tetrahedral mesh and four physiological parameters, the network outputs the displacement field that returns the myocardium to its unloaded state. On a held-out test set of 6,810 synthetic LV cases drawn from 20,700 finite-element simulations, the reported accuracy is a node-level DSC of 0.986 ± 0.023 and a Hausdorff distance of 0.083 ± 0.028 cm, with a mean node error of 0.028 ± 0.010 cm, at 0.023 seconds per case. This beats the inverse FE baseline on both accuracy and speed, and it also beats graph-convolution and point-cloud baselines. The authors attribute the data efficiency to a cycle-consistent bidirectional decoder: with about 200 supervised samples the model retains a DSC near 0.97, while removing the cycle term drops the same low-supervision setting to about 0.71. They present the method as the first learning-based approach to estimate unloaded cardiac geometry with explicit biophysical priors.

Load-bearing premise

The load-bearing premise is that the ground-truth unloaded geometry is the one produced by the paper's inverse-FE unloading-reloading pipeline, which the paper itself notes is typically under-constrained and can have non-unique solution, so a network trained to reproduce that pipeline may not recover the true zero-pressure state if the constitutive model or inverse solution is wrong.

Editorial extensions

If this is right

  • Inverse FE unloading can be replaced by one forward pass for meshes inside the trained distribution, cutting per-case cost from hours to roughly 0.02 seconds.
  • Cycle consistency enables a weak-supervision regime: about 200 labeled FE simulations are enough to keep node-level overlap near 0.97, reducing the cost of building training datasets.
  • Because the network is topology-agnostic and conditioned on pressure, stiffness, and fiber angles, it can process new meshes without template registration and interpolate to unseen physiological parameter values.
  • A fast, reliable zero-pressure reference state would make patient-specific stress and strain analysis and intervention planning practical in clinical time scales.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • An implication the paper leaves implicit is that the network is effectively a fast emulator of its own FE data-generation pipeline, so clinical value hinges on how faithfully that pipeline represents real myocardial tissue rather than on the reported synthetic accuracy alone.
  • A testable extension is to train the same cycle-consistent architecture on end-systolic or full-cycle meshes with active contraction parameters, since the loading and unloading decoders already form a bidirectional deformation model.
  • A practical deployment check is to run the trained network on image-derived patient meshes with pathologies; the current experiments cover healthy anatomical variation only, so behavior on scarred, dilated, or hypertrophic ventricles remains open.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. HeartUnloadNet is a graph-attention network that predicts the unloaded (zero-pressure) left-ventricular mesh from an end-diastolic mesh and physiological parameters (pressure, stiffness, helix angles). The model is trained on 20,700 paired FE simulations generated by an inverse unloading–reloading pipeline, uses a cycle-consistency loss for weak self-supervision, and is evaluated against GCN, PointNet++, and inverse FE baselines. The paper claims sub-millimeter accuracy (DSC 0.986, HD 0.083 cm), an inference time of 0.023 s (over 10^5 times faster than inverse FE), and that cycle consistency maintains 97% DSC with only ~200 training samples.

Significance. If the reported accuracy holds, this would be a valuable contribution: it is, to the authors' knowledge, the first deep-learning surrogate for unloaded cardiac geometry, it handles variable mesh topology without template registration, and it explicitly conditions on biophysical parameters. The release of code and the construction of a 20,700-case FE simulation dataset are also strengths, as is the systematic ablation study of architectural components. However, the central accuracy claims rest on a metric definition and on a data-efficiency claim that are internally inconsistent, and the ground truth is generated by an inverse FE pipeline whose non-uniqueness is acknowledged. These issues must be resolved before the performance claims can be accepted.

major comments (3)
  1. [Table 2, Table 3, Section 3.4] The definition of DSC as the percentage of nodes with Euclidean error below 0.1 mm (Table 2) is inconsistent with the reported MD and SD. For HeartUnloadNet, MD is 0.028 ± 0.010 cm (0.28 ± 0.10 mm) and SD is 0.013 cm (0.13 mm) (Table 3). A mean error of 0.28 mm with 98.6% of nodes below 0.1 mm would require a strongly non-Gaussian error distribution with extreme outliers; the reported SD alone does not support such a distribution without an explicit description. Moreover, the training protocol normalizes all mesh coordinates to zero mean and unit scale (Section 3.4), so it is unclear whether the 0.1 mm threshold is applied to de-normalized physical coordinates or to normalized coordinates. The authors should report the full error distribution (e.g., percentiles), clarify the coordinate frame used for DSC, and avoid implying that a high DSC independently establishes sub-millimeter accuracy when the threshold itself is 0.1 mm.
  2. [Section 4.2, Table 4, Abstract, Contribution 3] The claim that the model 'maintains a DSC of 97% even with as few as 200 training samples' is not supported by Table 4. Section 4.2 states that 3% labeled data corresponds to 'around 200 cases,' but 3% of the 13,890 training samples (42 shapes) is approximately 417 cases. The 1% row, corresponding to about 139 cases, yields DSC 0.619 ± 0.251 for the full model (A0). A 200-sample setting would fall between 1% and 3% supervision, with an expected DSC far below 0.97 based on the reported values. The abstract (200 samples), contribution 3 (209 training cases), and Section 4.2 (3%, ~200 cases) are mutually inconsistent. The authors must specify the exact number of labeled samples used in each supervision-ratio experiment and reconcile the text with the numbers in Table 4.
  3. [Section 3.3.1, Abstract, Section 4.1] The ground-truth unloaded geometry is produced by the authors' inverse FE unloading–reloading procedure, which the paper itself notes is 'typically under-constrained and can have non-unique solution.' The network therefore learns to reproduce the output of this particular inverse FE pipeline, not an independently validated zero-pressure state. The claim that HeartUnloadNet is 'significantly more accurate than traditional inverse FE solvers' (Section 4.1) compares the network against a different inverse FE baseline (Finsberg et al.), but both the network's training target and the baseline are outputs of inverse FE methods. This does not make the comparison circular in a formal sense, but it means the reported accuracy is relative to one chosen inverse FE solution, and the clinical-transferability claims are weaker than stated. The authors should explicitly acknowledge this limitation in the Limitations section and, if possible, validate on an independent forward-FE test or clinical data.
minor comments (5)
  1. [Section 3.3.1] In the sentence 'The only changed global parameters parameters allowed to vary', the word 'parameters' is duplicated; please proofread the manuscript throughout for similar typos.
  2. [Section 4.2] The description of weak supervision does not specify how the labeled subset is selected (e.g., random sampling, stratified by parameter) or whether the cycle-consistency loss is applied to all samples or only to unlabeled ones. This information is needed to interpret the supervision-ratio experiments.
  3. [Table 7] The LOVO table reports only a subset of the parameter values (three of six for pressure and stiffness). Please state why these particular values were chosen, or report all held-out values, so that the reader can assess the claimed interpolation capability.
  4. [Section 4.1] The inference time '0.023±0.002 s' in Table 3 differs from the abstract's '0.02 seconds' and the text's '~0.02 s'; please use consistent rounding or state the exact value in all places.
  5. [References] Reference [24] appears to be a self-citation to an arXiv preprint; please verify that this citation is appropriate and that all external references are correctly formatted.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity; the supervised regression is self-contained and the cycle-consistency term is a regularizer, not an independent label source. Minor non-load-bearing self-citations do not affect the derivation.

full rationale

HeartUnloadNet's central claim is a supervised surrogate: it is trained on paired (ED, unloaded) meshes generated by the authors' FE unloading-reloading pipeline, and tested on held-out synthetic meshes. The target labels are external to the network and are not constructed from the network's own outputs, so the accuracy numbers (DSC, HD, MD) are genuine measurements of fit to that FE-defined ground truth rather than tautologies. The cycle-consistency loss is an additional regularizer on the same paired data, L_total = ||hat G_u - G_u||^2 + lambda_cycle ||hat G_ED - tilde G_ED||^2, not an independent source of labels. The admitted non-uniqueness of the inverse FE unloading step (Section 3.3.1) is a limitation on the biophysical validity of the target, not a logical circularity in the network's derivation. The paper contains self-citations by the same group (refs [13], [24]) in background contexts only; none is load-bearing for the architecture, loss, or comparison. The reported inconsistency between '200 labeled samples' and the 3% supervision ratio in Table 4, and the nonstandard threshold-based definition of DSC in Table 2, are correctness and interpretability concerns that should be recomputed, but they do not make the derivation circular.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central claim rests on the FE simulation pipeline and the PCA shape model for defining both inputs and targets; the network itself introduces no new physical entities and fits no physical constants.

free parameters (2)
  • DSC distance threshold = 0.1 mm (claimed)
    Hand-chosen metric threshold that defines the reported DSC; it is inconsistent with reported MD and HD.
  • Cycle consistency weight = 0.2
    Hand-tuned loss weight; the paper does not report a sensitivity study.
assumptions (4)
  • domain assumption The myocardium follows a Fung-type transversely isotropic hyperelastic model with fixed anisotropic coefficients b_ff=29.9, b_xx=13.3, b_fx=26.6.
    Used to generate all FE ground truth and conditioned inputs; if the constitutive model is wrong, the network learns an incorrect inverse map. Eqs. 2-3.
  • domain assumption Helix angle varies linearly across the wall from endocardium to epicardium.
    Fiber orientation is a conditioning input; the assumed linear transmural variation may not hold in pathology. Section 3.1.
  • domain assumption The PCA shape model of 1,991 MESA subjects with weights truncated to [-2,2] covers clinically relevant LV shape variation.
    All training and test geometries are sampled from this model; pathological shapes are excluded. Section 3.3.1.
  • ad hoc to paper The unloaded geometry produced by the inverse FE unloading-reloading procedure is the correct target for the unloading task.
    The paper states the inverse FE problem is under-constrained with non-unique solutions, yet the network is trained to match the specific solution from this pipeline. Section 3.3.1.

how reviews work

0 comments
Cite this review

Pith. "Pith review of HeartUnloadNet: A Weakly-Supervised Cycle-Consistent Graph Network for Predicting Unloaded Cardiac Geometry from Diastolic States." pith.science (2026). https://pith.science/paper/GUFB322Y

@misc{pith2026250718677,
  author       = {Pith},
  title        = {Pith review of: HeartUnloadNet: A Weakly-Supervised Cycle-Consistent Graph Network for Predicting Unloaded Cardiac Geometry from Diastolic States},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GUFB322Y}},
  note         = {Machine review of arXiv:2507.18677}
}
read the original abstract

The unloaded cardiac geometry (i.e., the state of the heart devoid of luminal pressure) serves as a valuable zero-stress and zero-strain reference and is critical for personalized biomechanical modeling of cardiac function, to understand both healthy and diseased physiology and to predict the effects of cardiac interventions. However, estimating the unloaded geometry from clinical images remains a challenging task. Traditional approaches rely on inverse finite element (FE) solvers that require iterative optimization and are computationally expensive. In this work, we introduce HeartUnloadNet, a deep learning framework that predicts the unloaded left ventricular (LV) shape directly from the end diastolic (ED) mesh while explicitly incorporating biophysical priors. The network accepts a mesh of arbitrary size along with physiological parameters such as ED pressure, myocardial stiffness scale, and fiber helix orientation, and outputs the corresponding unloaded mesh. It adopts a graph attention architecture and employs a cycle-consistency strategy to enable bidirectional (loading and unloading) prediction, allowing for partial self-supervision that improves accuracy and reduces the need for large training datasets. Trained and tested on 20,700 FE simulations across diverse LV geometries and physiological conditions, HeartUnloadNet achieves sub-millimeter accuracy, with an average DSC of 0.986 and HD of 0.083 cm, while reducing inference time to just 0.02 seconds per case, over 10^5 times faster and significantly more accurate than traditional inverse FE solvers. Ablation studies confirm the effectiveness of the architecture. Notably, the cycle-consistent design enables the model to maintain a DSC of 97% even with as few as 200 training samples. This work thus presents a scalable and accurate surrogate for inverse FE solvers, supporting real-time clinical applications in the future.

Figures

Figures reproduced from arXiv: 2507.18677 by the authors.

Figure 1
Figure 1. [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Qualitative heatmap visualization of vertex-wise prediction errors on two representative test cases. Each row shows a different method (Inverse FE, PointNet++, GCN and HeartUnloadNet), and each column corresponds to one test case. Warmer colors indicate higher local displacement errors. 4.6. Advantages over PCA-Based and Inverse FE Approaches [PITH_FULL_IMAGE:figures/full_fig_p014_2.png] view at source ↗
Figure 3
Figure 3. Qualitative heatmap visualization comparing vertex-wise prediction errors of PCA and HeartUnloadNet for a representative test case. Compared to the ground truth, PCA exhibits a HD of 1.02 ± 0.07 cm, whereas HeartUnloadNet achieves 0.083 ± 0.028 cm. 5. Discussion 5.1. Novelty and Contribution The experimental results presented above validated not only the accuracy and efficiency of HeartUnloadNet, but also demonstrat… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

42 extracted references · 34 canonical work pages

  1. [1]

    Layer normalization

    Ba, J.L., Kiros, J.R., Hinton, G.E., 2016. Layer normalization. arXiv preprint arXiv:1607.06450

  2. [2]

    Finite element procedures

    Bathe, K.J., 1996. Finite element procedures

  3. [3]

    Large strain viscoelastic constitutive models

    Bonet, J., 2001. Large strain viscoelastic constitutive models. International Journal of Solids and Structures 38, 2953–2968

  4. [4]

    How attentive are graph attention networks? arXiv preprint arXiv:2105.14491

    Brody, S., Alon, U., Yahav, E., 2021. How attentive are graph attention networks? arXiv preprint arXiv:2105.14491

  5. [5]

    Application of feed forward and recurrent neural networks in simulation of left ventricular mechanics

    Dabiri, Y., Van der Velden, A., Sack, K.L., Choy, J.S., Guccione, J.M., Kassab, G.S., 2020. Application of feed forward and recurrent neural networks in simulation of left ventricular mechanics. Scientific reports 10, 22298

  6. [6]

    Emulation of cardiac mechanics using graph neural networks

    Dalton, D., Gao, H., Husmeier, D., 2022. Emulation of cardiac mechanics using graph neural networks. Computer Methods in Applied Mechanics and Engineering 401, 115645

  7. [7]

    Physics-informed graph neural network emulation of soft-tissue mechanics

    Dalton, D., Husmeier, D., Gao, H., 2023. Physics-informed graph neural network emulation of soft-tissue mechanics. Computer Methods in Applied Mechanics and Engineering 417, 116351

  8. [8]

    Model- based assessment of elastic material parameters in rheumatic heart disease patients and healthy subjects

    Familusi, M.A., Skatulla, S., Hussan, J.R., Aremu, O.O., Mutithu, D., Lumngwena, E.N., Gumedze, F.N., Ntusi, N.A., 2023. Model- based assessment of elastic material parameters in rheumatic heart disease patients and healthy subjects. Mathematical and Computational Applications 28, 106

Show all 42 references
  1. [9]

    Efficientestimationofpersonalizedbiventricular mechanicalfunctionemployinggradient-basedoptimization.Internationaljournalfornumericalmethodsinbiomedicalengineering34,e2982

    Finsberg,H.,Xi,C.,Tan,J.L.,Zhong,L.,Genet,M.,Sundnes,J.,Lee,L.C.,Wall,S.T.,2018. Efficientestimationofpersonalizedbiventricular mechanicalfunctionemployinggradient-basedoptimization.Internationaljournalfornumericalmethodsinbiomedicalengineering34,e2982

  2. [10]

    Modeling pathologies of diastolic and systolic heart failure

    Genet, M., Lee, L.C., Baillargeon, B., Guccione, J.M., Kuhl, E., 2016. Modeling pathologies of diastolic and systolic heart failure. Annals of biomedical engineering 44, 112–127

  3. [11]

    Gmsh: A 3-d finite element mesh generator with built-in pre-and post-processing facilities

    Geuzaine, C., Remacle, J.F., 2009. Gmsh: A 3-d finite element mesh generator with built-in pre-and post-processing facilities. International journal for numerical methods in engineering 79, 1309–1331

  4. [12]

    Gonzales, M.J., Sturgeon, G., Krishnamurthy, A., Hake, J., Jonas, R., Stark, P., Rappel, W.J., Narayan, S.M., Zhang, Y., Segars, W.P., et al.,

  5. [13]

    Myocardialbiomechanicaleffectsoffetalaorticvalvuloplasty

    Green,L.,Chan,W.X.,Tulzer,A.,Tulzer,G.,Yap,C.H.,2024. Myocardialbiomechanicaleffectsoffetalaorticvalvuloplasty. Biomechanics and Modeling in Mechanobiology 23, 1433–1448

  6. [14]

    Passive material properties of intact ventricular myocardium determined from a cylindrical model

    Guccione, J.M., McCulloch, A.D., Waldman, L., 1991. Passive material properties of intact ventricular myocardium determined from a cylindrical model

  7. [15]

    Inductive representation learning on large graphs

    Hamilton, W., Ying, Z., Leskovec, J., 2017. Inductive representation learning on large graphs. Advances in neural information processing systems 30

  8. [16]

    Constitutive modelling of passive myocardium: a structurally based framework for material characterization

    Holzapfel, G.A., Ogden, R.W., 2009. Constitutive modelling of passive myocardium: a structurally based framework for material characterization. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences 367, 3445–3475

  9. [17]

    Semi-supervised classification with graph convolutional networks

    Kipf, T.N., Welling, M., 2016. Semi-supervised classification with graph convolutional networks. arXiv preprint arXiv:1609.02907

  10. [18]

    Patient-specific models of cardiac biomechanics

    Krishnamurthy, A., Villongco, C.T., Chuang, J., Frank, L.R., Nigam, V., Belezzuoli, E., Stark, P., Krummen, D.E., Narayan, S., Omens, J.H., et al., 2013. Patient-specific models of cardiac biomechanics. Journal of computational physics 244, 4–21

  11. [19]

    Backpropagation applied to handwritten zip code recognition

    LeCun, Y., Boser, B., Denker, J.S., Henderson, D., Howard, R.E., Hubbard, W., Jackel, L.D., 1989. Backpropagation applied to handwritten zip code recognition. Neural computation 1, 541–551

  12. [20]

    Pytorch-fea: Autograd-enabled finite element analysis methods with applications for biomechanical analysis of human aorta

    Liang, L., Liu, M., Elefteriades, J., Sun, W., 2023. Pytorch-fea: Autograd-enabled finite element analysis methods with applications for biomechanical analysis of human aorta. Computer methods and programs in biomedicine 238, 107616

  13. [21]

    A machine learning approach as a surrogate of finite element analysis–based inverse method to estimate the zero-pressure geometry of human thoracic aorta

    Liang, L., Liu, M., Martin, C., Sun, W., 2018. A machine learning approach as a surrogate of finite element analysis–based inverse method to estimate the zero-pressure geometry of human thoracic aorta. International journal for numerical methods in biomedical engineering 34, e3103

  14. [22]

    Journal of computational physics 463, 111266

    Marx,L.,Niestrawska,J.A.,Gsell,M.A.,Caforio,F.,Plank,G.,Augustin,C.M.,2022.Robustandefficientfixed-pointalgorithmfortheinverse elastostatic problem to identify myocardial passive material parameters and the unloaded reference configuration. Journal of computational physics 463, 111266

  15. [23]

    Left ventricular shape variation in asymptomatic populations: the multi-ethnic study of atherosclerosis

    Medrano-Gracia, P., Cowan, B.R., Ambale-Venkatesh, B., Bluemke, D.A., Eng, J., Finn, J.P., Fonseca, C.G., Lima, J.A., Suinesiaputra, A., Young, A.A., 2014. Left ventricular shape variation in asymptomatic populations: the multi-ethnic study of atherosclerosis. Journal of Cardi...

  16. [24]

    Imc-pinn-fe: A physics-informed neural network for patient-specific left ventricular finite element modeling with image motion consistency and biomechanical parameter estimation

    Mu, S., Chan, W.X., Yap, C.H., 2025. Imc-pinn-fe: A physics-informed neural network for patient-specific left ventricular finite element modeling with image motion consistency and biomechanical parameter estimation. arXiv preprint arXiv:2506.20696

  17. [25]

    Effects of using the unloaded configuration in predicting the in vivo diastolic properties of the heart

    Nikou, A., Dorsey, S.M., McGarvey, J.R., Gorman III, J.H., Burdick, J.A., Pilla, J.J., Gorman, R.C., Wenk, J.F., 2016. Effects of using the unloaded configuration in predicting the in vivo diastolic properties of the heart. Computer methods in biomechanics and biomedical engin...

  18. [26]

    Learning mesh-based simulation with graph networks, in: International conference on learning representations

    Pfaff, T., Fortunato, M., Sanchez-Gonzalez, A., Battaglia, P., 2020. Learning mesh-based simulation with graph networks, in: International conference on learning representations

  19. [27]

    Pointnet++: Deep hierarchical feature learning on point sets in a metric space

    Qi, C.R., Yi, L., Su, H., Guibas, L.J., 2017. Pointnet++: Deep hierarchical feature learning on point sets in a metric space. arXiv preprint arXiv:1706.02413

  20. [28]

    An Introduction to Nonlinear Finite Element Analysis: with applications to heat transfer, fluid mechanics, and solid mechanics

    Reddy, J.N., 2015. An Introduction to Nonlinear Finite Element Analysis: with applications to heat transfer, fluid mechanics, and solid mechanics. Oxford university press

  21. [29]

    Highspatialresolutionmulti-organfiniteelementmodelingofventricular-arterialcoupling

    Shavik,S.M.,Jiang,Z.,Baek,S.,Lee,L.C.,2018. Highspatialresolutionmulti-organfiniteelementmodelingofventricular-arterialcoupling. Frontiers in physiology 9, 119

  22. [30]

    Heartsimsage: Attention-enhanced graph neural networks for accelerating cardiac mechanics modeling

    Shi, L., Chen, Y., Vedula, V., 2025. Heartsimsage: Attention-enhanced graph neural networks for accelerating cardiac mechanics modeling. arXiv preprint arXiv:2504.18968

  23. [31]

    Non-invasive in silico determination of ventricular wall pre-straining and characteristic cavity pressures

    Skatulla, S., Sansour, C., Familusi, M., Hussan, J., Ntusi, N., 2023. Non-invasive in silico determination of ventricular wall pre-straining and characteristic cavity pressures. arXiv preprint arXiv:2308.00461

  24. [32]

    Dropout: a simple way to prevent neural networks from overfitting

    Srivastava, N., Hinton, G., Krizhevsky, A., Sutskever, I., Salakhutdinov, R., 2014. Dropout: a simple way to prevent neural networks from overfitting. The journal of machine learning research 15, 1929–1958

  25. [33]

    Attention is all you need

    Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A.N., Kaiser, Ł., Polosukhin, I., 2017. Attention is all you need. Advances in neural information processing systems 30

  26. [34]

    Graphattentionnetworks

    Veličković,P.,Cucurull,G.,Casanova,A.,Romero,A.,Lio,P.,Bengio,Y.,2017. Graphattentionnetworks. arXivpreprintarXiv:1710.10903

  27. [35]

    Image-based predictive modeling of heart mechanics

    Wang, V., Nielsen, P., Nash, M., 2015. Image-based predictive modeling of heart mechanics. Annual review of biomedical engineering 17, 351–383

  28. [36]

    Efficientestimationofload-freeleftventriculargeometryandpassivemyocardialpropertiesusingprincipalcomponentanalysis

    Wang, Z.J., Wang, V.Y., Babarenda Gamage, T.P., Rajagopal, V., Cao, J.J., Nielsen, P.M., Bradley, C.P., Young, A.A., Nash, M.P., 2020. Efficientestimationofload-freeleftventriculargeometryandpassivemyocardialpropertiesusingprincipalcomponentanalysis. International journal for ...

  29. [37]

    Whitaker, J., Neji, R., Byrne, N., Puyol-Antón, E., Mukherjee, R.K., Williams, S.E., Chubb, H., O’Neill, L., Razeghi, O., Connolly, A., et al.,

  30. [38]

    Non-linear finite element analysis of solids and structures, volume 1: Essentials, ma crisfield, john wiley

    Wriggers, P., 1994. Non-linear finite element analysis of solids and structures, volume 1: Essentials, ma crisfield, john wiley. isbn: 0-471- 92956-5

  31. [39]

    Graph transformer networks

    Yun, S., Jeong, M., Kim, R., Kang, J., Kim, H.J., 2019. Graph transformer networks. Advances in neural information processing systems 32

  32. [40]

    Unpaired image-to-image translation using cycle-consistent adversarial networks, in: Proceedings of the IEEE international conference on computer vision, pp

    Zhu, J.Y., Park, T., Isola, P., Efros, A.A., 2017. Unpaired image-to-image translation using cycle-consistent adversarial networks, in: Proceedings of the IEEE international conference on computer vision, pp. 2223–2232. S.Mu et al.: Preprint submitted to Elsevier Page 17 of 17

  33. [2013]

    Medical image analysis 17, 525–537

    Athree-dimensionalfiniteelementmodelofhumanatrialanatomy:Newmethodsforcubichermitemesheswithextraordinaryvertices. Medical image analysis 17, 525–537. S.Mu et al.: Preprint submitted to Elsevier Page 16 of 17 HeartUnloadNet

  34. [2019]

    Journal of Cardiovascular Magnetic Resonance 21, 62

    Improved co-registration of ex-vivo and in-vivo cardiovascular magnetic resonance images using heart-specific flexible 3d printed acrylic scaffold combined with non-rigid registration. Journal of Cardiovascular Magnetic Resonance 21, 62

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.