Pith. sign in

REVIEW 4 major objections 6 minor 15 references

Synthetic heart-valve shapes can be generated by sampling a Bayesian posterior built from a Gaussian-mixture prior and a classifier-based validity likelihood, outperforming PCA-based shape models when data are scarce.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · deepseek-v4-flash

2026-08-03 01:22 UTC pith:ZOWFOUKY

load-bearing objection Sensible Bayesian re-framing of SSM shape generation with a clean 2D verification, but the aortic-valve low-data claim is self-referential until the validity classifier is independently validated. the 4 major comments →

arxiv 2607.28914 v1 pith:ZOWFOUKY submitted 2026-07-31 cs.CE

Bayesian Posterior Sampling for Synthetic Shape Generation of Heart Valves

Vijay Dubey , Sumedh Seetharam , Nikos Manthatis , Collin E. Haese , L. River Spencer , Jakob Christoph Voran , Marnie Goldmann , Felix Kreidel
show 4 more authors
Issam Moussa Alison Pouch Manuel K. Rausch Jan Fuhg
This is my paper
classification cs.CE
keywords statistical shape modelsBayesian posterior samplingheart valve shape generationGaussian mixture modelsNo-U-Turn samplerproper orthogonal decompositiontricuspid valvedata augmentation
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

This paper claims that the right way to generate synthetic heart-valve shapes from small cohorts is to sample from a Bayesian posterior over a compact shape-coefficient space, rather than to sample coefficients from a PCA-based statistical shape model. The prior is a Gaussian mixture fitted to real valve shapes; the likelihood is a classifier trained to tell physically valid reconstructions from defective ones; the No-U-Turn sampler draws from the resulting posterior. On a controlled model problem and on parametrically generated aortic valves, the method captures multimodal structure and yields higher fractions of valid shapes than PCA-based baselines, even with as few as ten training shapes. Applied to ten segmented human tricuspid valves from 3D ultrasound, it produces synthetic but physiologically plausible valves that can feed downstream image-mask generation and finite-element simulation.

Core claim

The central claim is that valve-shape generation should be posed as posterior sampling conditioned on validity: π(shape|good) ∝ π(good|shape)π(shape). Valves become corresponding NURBS control-point vectors, compressed by proper orthogonal decomposition into a low-dimensional coefficient space; a Gaussian mixture prior is fitted to real shapes; a multilayer-perceptron classifier supplies the likelihood from reconstructed shapes labeled good or bad by rule or visual inspection; and the No-U-Turn sampler draws from the posterior. The paper reports this captures multimodality, respects validity boundaries, and outperforms PCA-based statistical shape models in the low-data regime (N=10), enablin

What carries the argument

The load-bearing object is the posterior π(a|good) ∝ π_C(good|a) π_G(a) in truncated POD coefficient space. π_G is a Gaussian mixture model with a data-driven number of modes; π_C is a multilayer-perceptron classifier that, for valve problems, operates on an eight-dimensional feature vector computed from reconstructed surfaces (self-intersection energy, tangent orthogonality, normal deviation, worst aspect ratio, area variability, symmetry, and fold indicators). NUTS provides gradient-based exploration of this posterior, which is what allows conditional sampling without rejection inefficiency.

Load-bearing premise

The classifier trained on visually inspected generated shapes is a complete and transferable definition of physiological validity; if its boundary is biased by the generator's sampling density or fails on posterior samples, the method concentrates on shapes that are not truly valid.

What would settle it

Generate 500 synthetic tricuspid valves with the proposed sampler and have clinicians who did not label the training data independently judge plausibility; if the clinician-approved fraction is not clearly higher than a PCA-based SSM baseline, or disagrees substantially with the classifier's labels, the central claim fails. Similarly, an independent geometric oracle—for example, running finite-element coaptation or mesh-quality checks on generated shapes—could reveal defects the eight-feature classifier misses.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Synthetic valve shapes generated from as few as ten segmented tricuspid valves can be paired with synthetic ultrasound rendering to bootstrap training sets for autosegmentation models.
  • The generated shapes plug directly into finite-element valve models, enabling cohort studies of valve mechanics over synthetic anatomical variation.
  • The conditioning mechanism is general: any criterion that can be labeled and learned by a classifier—morphological type, clinical plausibility—can replace or join geometric validity.
  • The aortic-valve experiments indicate diminishing returns around N≈100 segmented shapes, giving a practical target for dataset collection.
  • Compared with a generator-rejector approach, NUTS maintains validity while exploring a broader region of shape space, which matters as coefficient-space dimension grows.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the classifier's notion of 'good' is incomplete—e.g., misses defect modes not expressible in the eight geometric features—the posterior may certify shapes that are not genuinely valid; an independent clinical or oracle labeling study on the generated cohort would test this.
  • At N=10 the Gaussian mixture prior effectively collapses toward a single mode, so the classifier term is doing most of the conditioning work; ablating the prior and classifier separately would reveal which component carries low-data performance.
  • The same posterior formulation should extend to conditional generation on morphological type once datasets contain enough examples per type; the paper notes four tricuspid types but does not attempt it.
  • Because NUTS requires differentiability of both prior and likelihood, non-differentiable validity criteria such as full finite-element coaptation checks cannot enter directly; a differentiable surrogate likelihood would be a natural next step.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 6 minor

Summary. The paper proposes a Bayesian posterior sampling framework for synthetic heart-valve shape generation. Shapes are represented as NURBS surfaces and mapped to a truncated POD coefficient space. A Gaussian mixture prior is fitted to the real coefficient vectors, and a multilayer-perceptron classifier C, trained on generator samples labeled by a prescribed boundary or by visual inspection, defines a validity likelihood. Samples are drawn from the posterior using the No-U-Turn sampler (NUTS). The method is verified on a two-dimensional mixture model with an analytical decision boundary, validated on parametrically generated aortic valves, and applied to ten 3D TEE-derived tricuspid valves. The authors claim that the method captures multimodality, respects validity boundaries, and outperforms PCA-based SSMs in low-data regimes, with downstream demonstrations in synthetic ultrasound generation and finite-element valve mechanics.

Significance. If the central claims hold, the paper offers a principled solution to a recognized limitation of PCA-based SSMs—namely, their inability to handle multimodal, non-Gaussian coefficient distributions or to enforce validity constraints during generation. The 2D model problem is a genuine ground-truth verification: the sampling mechanism demonstrably recovers multiple modes and respects an analytical boundary. The GMM-prior + classifier-likelihood decomposition is general and could be applied to other anatomical structures and conditioning criteria. The paper is also honest about several limitations, including the scalar validity label source, the energy-threshold choice, and the lack of inter-observer variability analysis. However, the load-bearing evidence for the headline 'outperforms PCA-based SSMs in low-data regimes' claim rests on the aortic-valve valid-fraction metric, which is evaluated with the same classifier that constitutes the NUTS likelihood. Until that circularity is broken by an independent validity oracle or by a posterior-sample-specific classifier validation, the superiority claim for anatomical shapes remains conditional.

major comments (4)
  1. [§4.2.1, Eq. (37) and §2.4.3, Eq. (19)] The aortic-valve valid fraction f_valid is computed using the classifier C (Section 4.2.1), which is the exact same classifier whose output defines the NUTS likelihood in Eq. (19). Thus, in Fig. 5(b), NUTS is being rewarded on its own objective while the SSM baseline never sees that score. This makes the central low-data superiority claim self-referential. The model problem avoids this because labels come from an analytical boundary, but the aortic-valve problem is the primary evidence for the 'outperforms PCA-based SSMs' claim. I recommend either (i) validating C against an independent oracle on NUTS-generated samples, e.g., expert or clinician labels of reconstructed meshes, or (ii) reporting additional validity metrics derived from geometric invariants not used in Eq. (23). Without this, the aortic comparison cannot support the stated conclusion.
  2. [§3.2 and Fig. 5] All results in Fig. 5 appear to be single-run point estimates. The subset construction in Section 3.2 draws one random subset per size, and no repeated realizations, bootstrap, or confidence intervals are reported. For the low-data regime (N=10), the NUTS sampler retains only 500 pooled samples (Appendix B, Table B.2), so the metrics in Fig. 5(b) may be highly variable. I suggest repeating the subset construction over multiple seeds (e.g., 10-20 random draws) and reporting mean and spread, or at least showing error bars for the aortic-valve panels. Without such variation, it is difficult to assess whether the claimed advantage at N=10 is robust or an artifact of a single data draw.
  3. [§2.4.1, Appendix B] The manuscript does not report MCMC convergence diagnostics (e.g., R-hat, effective sample size, trace plots) for the NUTS chains. Given that the aortic- and tricuspid-valve problems retain only 100 draws per chain across 5 chains (500 total samples) in a coefficient space that can be up to 50-dimensional (Section 5), the fidelity and diversity metrics in Fig. 5 may be based on poorly converged or strongly autocorrelated samples. Please report convergence statistics and, if necessary, increase the number of retained draws or thinning. This is standard practice for Bayesian posterior sampling and is directly relevant to the quantitative comparisons.
  4. [§2.4.3, Eq. (23), and §5] The classifier C is trained on samples from the generator G and labeled by visual inspection (Section 2.4.3). The paper acknowledges in Section 5 that inter-observer variability was not examined. However, this point is load-bearing because the method's 'physiologically plausible' guarantee is only as strong as C's transferability from prior-generated to posterior-generated shapes. The eight features in Eq. (23) may not capture all defect modes, and the classifier boundary may be biased by G's sampling density. I suggest a targeted experiment: train C on prior samples, then have independent experts label a set of NUTS-generated shapes (or a held-out set of real shapes) and compare the classifier's predictions against expert labels. Reporting per-class accuracy, calibration, and inter-observer agreement would substantially strengthen the validity claim.
minor comments (6)
  1. [Abstract and §4.5] The abstract states that the approach 'bootstraps building image-mask datasets more efficiently than PCA-based SSMs.' However, Section 4.5 only demonstrates a single synthetic ultrasound image and a finite-element simulation; no quantitative comparison of downstream segmentation performance or dataset-augmentation efficiency is provided. Please either temper this claim or add an experiment comparing segmentation performance when training on real versus synthetic-augmented data.
  2. [§2.4.3, Eq. (24)] The tangent-point energy exponent α is set to 2 by default, but no sensitivity analysis is provided. Since α controls the sharpness of the self-intersection penalty and thereby influences the classifier's feature distribution, a brief discussion or one sensitivity check would be helpful.
  3. [§3.1.2] The aortic-valve parametric generator uses a 'common standard deviation' chosen via visual inspection. This choice affects the diversity of the ground-truth shapes and therefore the difficulty of the validation problem. Please disclose the exact standard deviation value and, ideally, provide a sensitivity analysis or at least state how the value was selected.
  4. [§4.4] The tricuspid-valve application is presented only qualitatively (Fig. 6). While the small dataset size (N=10) limits quantitative evaluation, the paper could still report simple metrics on the generated tricuspid shapes, such as valid fraction under the classifier, or geometric feature distributions, to give the reader more than a visual impression. This would also help contextualize the 'limited diversity' noted by the authors.
  5. [Fig. 5] The panels in Fig. 5 would be easier to interpret if the axes ranges and subset sizes (e.g., 10, 50, 100, 350) were labeled directly, rather than relying on the reader to infer them from the text. Additionally, the ordering 'NUTS≥SSM≥GR' in the Wasserstein row is described in the text; adding this ordering to the figure or a table would improve clarity.
  6. [Section 5] The claim that NUTS 'samples farther into the underlying distribution' is based on Wasserstein distance and mean/covariance deviations. Since these metrics are computed against the same reference set used for training (N=350 for aortic), it would be useful to discuss potential bias from using the reference set that is also the largest training subset.

Circularity Check

1 steps flagged

Aortic valid-fraction metric uses the same classifier that forms the NUTS likelihood; low-data superiority claim is self-referential.

specific steps
  1. fitted input called prediction [Eq. (19), Section 2.4; Section 4.2.1, Eq. (37), Fig. 5b]
    "π(a(s)1 | c(a(s)1)=1) ∝ πC(c(a(s)1)=1 | a(s)1) πG(a(s)1) ... For the aortic-valve problem, no such analytical boundary is available; therefore, we use the trained classifier C to estimate whether each generated shape lies in the good region."

    The NUTS posterior target is exactly the classifier likelihood πC(c=1|a) times the prior. The aortic valid fraction f_valid (Eq. 37) then counts the fraction of NUTS samples for which this same classifier returns c=1. Thus the aortic f_valid curve in Fig. 5b measures how well NUTS satisfies its own likelihood, while the SSM baseline is never scored by that likelihood. The classifier is also trained on samples drawn from the generative prior G and labeled by visual inspection, so C encodes G's sampled manifold rather than an independent validity oracle. The model problem avoids this because it uses the prescribed decision boundary, but the abstract's 'outperforms PCA-based SSMs in low-data regimes' claim is carried by the aortic experiment, making the central empirical claim self-referentia

full rationale

The paper's 2D model problem is a genuine verification: labels come from the prescribed decision boundary, so f_valid, density, coverage, and Wasserstein metrics there are independent of the fitted classifier. The aortic-valve problem, however, is the primary evidence for the low-data superiority claim. There, the validity likelihood in Eq. (19) and the f_valid metric in Eq. (37) are the same trained classifier C. NUTS samples from a posterior that contains C as its likelihood, so penalizing or rewarding NUTS with C is partially optimizing its own objective; SSM never sees C and therefore cannot be compared on that score as an independent measure of physiological plausibility. The paper openly acknowledges that aortic labels come from visual inspection and that inter-observer variability was not studied, which reinforces that C has no external oracle. The proposed method is plausible and the framework is not circular in its construction, but the headline anatomical low-data claim is not independently established by the aortic valid-fraction metric. Score 6 reflects one load-bearing self-referential evaluation, while the model-problem verification and the other metrics provide some independent content.

Axiom & Free-Parameter Ledger

5 free parameters · 7 axioms · 0 invented entities

No new physical entities are postulated. The eight-dimensional quality feature vector and the classifier's good/bad boundary are constructed latent quantities, but they are functions of the geometry, not new ontology. The load-bearing assumptions are about transfer of the learned validity function and the completeness of the feature set.

free parameters (5)
  • POD energy threshold E_T = 0.95
    Truncation threshold in Eq. (14)–(15); set to 95% by convention, acknowledged in Section 5 as not investigated.
  • GMM upper bound on mixture components = 10
    Upper limit set by hand (Section 2.4.2); actual mode count selected by scikit-learn defaults and data.
  • Aortic parameter std dev = not reported
    Univariate Gaussian std dev for parametric ground truth "chosen via visual inspection of the resulting generated shapes" (Section 3.1.2).
  • Tangent point energy exponent alpha = 2
    Default exponent in R_tpe feature, Eq. (24); a modeling choice for the self-intersection indicator.
  • Manual good/bad labels for classifier = n/a (expert process)
    Valve classifier labels come from visual inspection of G-generated meshes (Section 2.4.3); these labels are the ground truth for the validity likelihood.
axioms (7)
  • domain assumption ValveFit NURBS fitting yields corresponding, analysis-suitable surfaces across subjects.
    Section 2.1; the whole pipeline inherits ValveFit's correspondence and regularization assumptions.
  • domain assumption 95% POD energy truncation preserves all physiologically relevant shape variation.
    Eq. (14)–(15); acknowledged as uninvestigated in Section 5.
  • ad hoc to paper The classifier trained on generator-G samples generalizes to the posterior region.
    Section 2.4.3; load-bearing: if the boundary learned from G's samples is biased, posterior samples inherit the bias.
  • ad hoc to paper The eight geometric quality features (Eq. 23) capture all relevant defect modes.
    Section 2.4.3; defects absent from the features are invisible to the likelihood.
  • domain assumption Visual inspection reliably determines physiological plausibility.
    Section 2.4.3; no inter-observer validation performed (acknowledged in Section 5).
  • domain assumption Manual 3D TEE segmentations are accurate shape ground truth.
    Section 3.1.3; "no formal analysis of potential systematic bias" stated.
  • standard math NUTS/HMC converges to the target posterior for the learned GMM × classifier density.
    Section 2.4.1; relies on smoothness/differentiability of both terms.

pith-pipeline@v1.3.0-daily-deepseek · 210 in / 17136 out tokens · 227650 ms · 2026-08-03T01:22:41.381251+00:00 · methodology

0 comments
read the original abstract

Statistical shape models (SSMs) for heart valves commonly rely on principal component analysis (PCA). They are used to support downstream tasks, including \textit{in silico} modeling, morphological analysis, and interventional planning. However, PCA-based SSMs lack a mechanism for conditional shape generation, i.e., they can create non-physical shapes and perform poorly in low-data regimes (<20 shapes). To overcome these problems, we propose instead a Bayesian posterior sampling framework to generate valve shapes from a posterior estimate. The prior relies on a Gaussian mixture model with data-driven mixture modes. The likelihood estimate is obtained through a classifier trained to distinguish valid from invalid regions in the compact proper orthogonal decomposition (POD) coefficient space. We verify the framework on a model problem and validate it on parametrically constructed aortic valve datasets. Thereby, we demonstrate that our method captures multiple modes, respects decision boundaries in shape space, and outperforms PCA-based SSMs in low-data regimes. We also characterize the framework's performance as a function of dataset size, identifying where diminishing returns arise for the proposed generative shape model. Finally, we apply the framework to a cohort of ten three-dimensional transesophageal echocardiography images of adult human tricuspid valves. We first segment images to extract shapes, then generate a set of physiologically plausible new shapes. We demonstrate downstream applications for both valves, including \textit{in silico} modeling of valve mechanics and synthetic image-mask creation to augment limited datasets. The proposed approach bootstraps building image-mask datasets more efficiently than PCA-based SSMs. Although demonstrated only for the aortic and tricuspid valves, the methodology is broadly applicable to all valves.

Figures

Figures reproduced from arXiv: 2607.28914 by Alison Pouch, Collin E. Haese, Felix Kreidel, Issam Moussa, Jakob Christoph Voran, Jan Fuhg, L. River Spencer, Manuel K. Rausch, Marnie Goldmann, Nikos Manthatis, Sumedh Seetharam, Vijay Dubey.

Figure 1
Figure 1. Figure 1: Various priors fit to the generated test data, where the top row shows points in [PITH_FULL_IMAGE:figures/full_fig_p008_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Various copula prior fits to the generated test data, where the top row shows points in [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Various problems tested using the current framework. [PITH_FULL_IMAGE:figures/full_fig_p012_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Synthetic data generated using SSM and NUTS given the same starting real dataset for the model problem. We varied the number of [PITH_FULL_IMAGE:figures/full_fig_p014_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Various performance metrics for (a) the model problem (2D) and (b) the Aortic Valve problem (3D) as a function of the size of the starting subset drawn from the target distribution. SSM is the statisti￾cal shape method, GR is the generator-rejector method, and NUTS is the current proposed sampling strategy (Section 2.5). row of [PITH_FULL_IMAGE:figures/full_fig_p015_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Real and synthetic (NUTS) tricuspid valves. [PITH_FULL_IMAGE:figures/full_fig_p017_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: We generated synthetic shapes for the aortic and the tricuspid valve [PITH_FULL_IMAGE:figures/full_fig_p017_7.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

15 extracted references · 2 canonical work pages · 1 internal anchor

  1. [5]

    Data in Brief 39, 107664

    An in-silico benchmark for the tricuspid heart valve–geometry, finite element mesh, abaqus simula- tion, and result data set. Data in Brief 39, 107664. Lin, F., Zakeri, A., Xue, Y ., MacRaild, M., Dou, H., Zhou, Z., Zou, Z., Sarrami-Foroushani, A., Duan, J., Frangi, A.F., 2026. From pix- els to polygons: A survey of deep learning approaches for medical im...

    arXiv 2026

  2. [7]

    Medical Image Analysis 42, 60–

    A survey on deep learning in medical image analysis. Medical Image Analysis 42, 60–

  3. [10]

    IEEE Access 12, 5295–5308

    A Deep Learning-Based Fully Automated Pipeline for Regurgitant Mitral Valve Anatomy Analysis From 3D Echocardiography. IEEE Access 12, 5295–5308. URL:https://ieeexplore. ieee.org/abstract/document/10380557, doi:10.1109/ACCESS.2024.3349698. Munsell, B.C., Dalal, P., Wang, S., 2008. Evaluating Shape Correspondence for Statistical Shape Analy- 21 sis: A Benc...

    arXiv 2024

  4. [12]

    Position: All Current Generative Fidelity and Diversity Metrics are Flawed

    Image Segmentation and Modeling of the Pediatric Tricuspid Valve in Hypoplastic Left Heart Syndrome, in: Pop, M., Wright, G.A. (Eds.), Func- tional Imaging and Modelling of the Heart. Springer International Publishing, Cham. volume 10263, pp. 95–105. URL:http://link.springer.com/ 10.1007/978-3-319-59448-4_10, doi:10.1007/ 978-3-319-59448-4_10. series Titl...

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2505.22450 2019

  5. [32]

    1016/B978-0-12-810493-4.00002-X

    URL:https://linkinghub.elsevier.com/ retrieve/pii/B978012810493400002X, doi:10. 1016/B978-0-12-810493-4.00002-X. Litjens, G., Kooi, T., Bejnordi, B.E., Setio, A.A.A., Ciompi, F., Ghafoorian, M., van der Laak, J.A.W.M., van Ginneken, B., Sánchez, C.I.,

  6. [88]

    Maas, S.A., Ellis, B.J., Ateshian, G.A., Weiss, J.A.,

    URL:https://www.sciencedirect.com/ science/article/pii/S1361841517301135, doi:10.1016/j.media.2017.07.005. Maas, S.A., Ellis, B.J., Ateshian, G.A., Weiss, J.A.,

    doi:10.1016/j.media.2017.07.005 2017

  7. [180]

    Kunzelman, K.S., Quick, D.W., Cochran, R.P., 1998

    URL:https://www.sciencedirect.com/ 20 science/article/pii/S1361841516301128, doi:10.1016/j.media.2016.06.035. Kunzelman, K.S., Quick, D.W., Cochran, R.P., 1998. Altered collagen concentration in mitral valve leaflets: biochemical and finite element analysis. The Annals of thoracic surgery 66, S198–S205. Lasso, A., Herz, C., Nam, H., Cianciulli, A., Pieper...

    doi:10.1016/j.media.2016.06.035 2016

  8. [2012]

    Journal of Biomechanical Engineering 134, 011005

    FEBio: finite elements for biomechan- ics. Journal of Biomechanical Engineering 134, 011005. URL:https://doi.org/10.1115/1. 4005694, doi:10.1115/1.4005694. MacKay, D.J.C., 2019. Information theory, inference, and learning algorithms. 22nd printing ed., Cam- bridge University Press, Cambridge. Mathur, M., Meador, W.D., Malinowski, M., Jazwiec, T., Timek, T...

    arXiv 2019

  9. [2017]

    Journal of the American Statisti- cal Association 112, 859–877

    Variational Inference: A Review for Statisticians. Journal of the American Statisti- cal Association 112, 859–877. URL:https: //doi.org/10.1080/01621459.2017.1285773, doi:10.1080/01621459.2017.1285773. _eprint: https://doi.org/10.1080/01621459.2017.1285773. Bradbury, J., Frostig, R., Hawkins, P., Johnson, M.J., Katariya, Y ., Leary, C., Maclaurin, D., Nec...

    arXiv 2017

  10. [2018]

    Psychonomic Bulletin & Review 25, 143–154

    A simple introduction to Markov Chain Monte–Carlo sampling. Psychonomic Bulletin & Review 25, 143–154. URL:http://link. springer.com/10.3758/s13423-016-1015-8, doi:10.3758/s13423-016-1015-8. Wu, W., Ching, S., Maas, S.A., Lasso, A., Sabin, P., Weiss, J.A., Jolley, M.A., 2022. A Computational Framework for Atrioventricular Valve Modeling Us- ing Open-Sourc...

    arXiv 2022

  11. [2021]

    font-variant:small- caps

    Segmentation of Tricuspid Valve Leaflets From Transthoracic 3D Echocardiograms of Children With Hypoplastic Left Heart Syndrome Using Deep Learning. Frontiers in Cardiovascular Medicine 8, 735587. URL:https://www.frontiersin.org/ articles/10.3389/fcvm.2021.735587/full, doi:10.3389/fcvm.2021.735587. Hitchcock, D.B., 2003. A History of the Metropo- lis–Hast...

    arXiv 2021

  12. [2024]

    Journal of the American Society of Echocardiography 37, 259–267

    Euclidean and Shape-Based Analysis of the Dynamic Mitral Annulus in Children using a Novel Open-Source Framework. Journal of the American Society of Echocardiography 37, 259–267. URL:https://onlinejase.com/ article/S0894-7317(23)00594-1/abstract, doi:10.1016/j.echo.2023.11.011. Anderson, A.E., Ellis, B.J., Weiss, J.A., 2007. Verification, validation and s...

    arXiv 2023

  13. [2026]

    URL:https://www.biorxiv.org/ content/10.64898/2026.07.13.736182v1, doi:10.64898/2026.07.13.736182

    Rt3DE-based finite element analysis of func- tional tricuspid regurgitation and RV free wall ap- proximation. URL:https://www.biorxiv.org/ content/10.64898/2026.07.13.736182v1, doi:10.64898/2026.07.13.736182. iSSN: 2692- 8205 Pages: 2026.07.13.736182 Section: New Results. Van Ravenzwaaij, D., Cassey, P., Brown, S.D.,

    work page doi:10.64898/2026.07.13.736182v1 2026

  14. [2039]

    URL:https://ieeexplore.ieee.org/ document/4415270, doi:10.1109/TPAMI.2007. 70841. Naeem, M.F., Oh, S.J., Uh, Y ., Choi, Y ., Yoo, J., 2020. Reliable Fidelity and Diversity Metrics for Gen- erative Models. URL:http://arxiv.org/abs/ 2002.09797, doi:10.48550/arXiv.2002.09797. arXiv:2002.09797 [cs]. Park, J.J., Florence, P., Straub, J., Newcombe, R., Lovegrov...

    arXiv 2007

  15. [5782]

    2025.11357130

    URL:https://ieeexplore.ieee.org/ document/11357130, doi:10.1109/BIBM66473. 2025.11357130. iSSN: 2156-1133. Zhao, A., Balakrishnan, G., Durand, F., Guttag, J.V ., Dalca, A.V ., 2019. Data augmentation using learned transformations for one-shot medical image segmen- tation, in: Proceedings of the IEEE/CVF Confer- ence on Computer Vision and Pattern Recognit...

    arXiv 2025