Pith. sign in

REVIEW 3 major objections 5 minor 51 references

Bayesian finite element regression for vascular flow reconstruction with quantified uncertainty

T0 review · 3 major / 5 minor · reviewed 2026-08-01 · deepseek-v4-flash

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

Pith's one-line read This paper claims that a Bayesian finite-element regression method, driven by a maximum-entropy physics prior and a Gaussian noise model, reconstructs steady three-dimensional velocity and pressure fields with quantified uncertainty from no

desk verdict A well-derived training-free Bayesian FEM reconstruction framework that is genuinely useful within its stated scope; the main gap is that the evaluation is synthetic and inverse-crime-adjacent, so the accuracy and UQ claims are not yet tested against mismatch. read the letter →

arxiv 2607.20224 v1 pith:AVPMOMM5 submitted 2026-07-22 physics.flu-dyn

Cem Gormezano , Shawn Shadden This is my paper · ORCID
classification physics.flu-dyn MSC 65N2165N3076D0562F15
keywords CardiovascularflowreconstructionBayesianinverseproblemsFiniteelementmethodUncertaintyquantificationMaximum-entropypriorsPhase-contrastMRIWallshearstressNavier-Stokesequations
topics Navier-Stokes Regularity
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 sets out to solve a clinical inverse problem: from noisy, low-resolution velocity images of blood flow, recover full three-dimensional velocity and pressure fields, plus trustworthy error bars, without any training data or repeated CFD solves. The proposed finite element regression treats the unknown fields as Taylor-Hood finite element coefficients, encodes smoothness and the Navier-Stokes equations as maximum-entropy priors, and obtains the maximum a posteriori estimate by solving one sparse nonlinear least-squares problem; a Laplace approximation then turns the posterior Hessian into uncertainty bounds for fields and derived quantities like flow rate, pressure drop, and wall shear stress. If correct, the method would give clinicians super-resolved hemodynamics from PC-MRI with uncertainty, at a fraction of PINN cost, and with region-of-interest wall shear stress errors several times smaller than either tricubic interpolation or a PINN baseline.

What carries the argument

The central object is the maximum-entropy prior placed on Taylor-Hood (continuous piecewise-quadratic velocity, piecewise-linear pressure) finite element degrees of freedom. Navier-Stokes residuals are projected onto the test space and their expected norms enter as constraints, yielding a Gaussian prior whose precision is a weighted sum of mass/stiffness matrices and projected momentum/continuity operators. The MAP estimate solves the resulting sparse nonlinear least-squares problem after pressure is eliminated analytically, and the Laplace approximation gives the posterior covariance as the inverse Hessian at the MAP, which propagates analytically to linear functionals such as flow rate, av

What would settle it

Generate a synthetic truth with a fine spectral or higher-order finite-element discretization (or with pulsatile inlet conditions), downsample and add realistic PC-MRI noise with spatially varying variance and imaging artifacts, and check whether the method's region-of-interest velocity and wall shear stress errors remain in the reported low range and whether empirical coverage stays near 95%; if errors inflate or coverage drops, the claim is bounded to the same-model setup.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central claim is that the posterior formed from a Gaussian likelihood for noisy velocity observations and a maximum-entropy prior constrained by boundedness, smoothness, incompressibility, and momentum balance has a MAP estimate that accurately reconstructs steady velocity and pressure in patient-specific aneurysms and coarctation, and a Laplace covariance that is conservative (roughly 99% of nodal values inside nominal 95% intervals). The authors demonstrate this on synthetic data for three geometries across SNR 2.5–10 and resolutions 0.5–2.5 mm, reporting region-of-interest WSS errors near or below a few percent for the proposed method, substantially lower tha

Load-bearing premise

The evaluation assumes the observed data are actually produced by steady Navier-Stokes solutions drawn from the same finite-element family the method uses, corrupted by independent Gaussian noise of known variance; if real PC-MRI noise, pulsatility, or a truth field outside that space breaks this alignment, the reported accuracy and uncertainty calibration could change.

Editorial extensions

If this is right

  • Steady vascular flow fields can be reconstructed from PC-MRI-like velocity data without any training set, removing the need to match a database of similar geometries.
  • Wall shear stress, the quantity most sensitive to near-wall resolution, can be recovered with a few percent error in diseased regions, where clinical decisions hinge.
  • Because the Laplace covariance is derived from the same Hessian used for optimization, uncertainty bands for flow rate, pressure drop, and WSS are essentially free once the MAP is found.
  • The weakly divergence-free variant converges in roughly half the iterations, making the physics-constrained reconstruction practical at about 15 minutes per case on a single CPU.
  • Pressure fields, which are not directly measured, are inferred jointly with velocity, giving access to pressure gradients and pressure drop without a separate pressure-Poisson solve.

Reading between the lines

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

  • The reported coverage being near 99% on nominal 95% intervals suggests the Laplace approximation overestimates variance; a tighter posterior that accounts for the skewness introduced by the advection term might give better-calibrated bands.
  • The framework's success likely depends on the data being generated by steady Navier-Stokes in the same Taylor-Hood discretization family; on genuinely pulsatile flow or on a truth field outside the finite element space, the physics prior becomes a model error rather than the true generative process, and the reported ranking could change.
  • A natural testable extension is to apply the same max-entropy prior idea to the Fourier coefficients of time-periodic flow, which the authors name as future work; the Laplace covariance machinery would carry over directly.
  • One could probe the informativeness of the physics prior by removing the momentum term (setting its weight to zero) and checking how much of the wall shear stress accuracy is due to the Navier-Stokes prior versus the smoothness prior plus data alone.
Share X Bluesky LinkedIn Reddit HN

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. The paper develops a Bayesian finite element regression (FER) framework for reconstructing steady, three-dimensional velocity and pressure fields from noisy, under-resolved velocity measurements. Fields are represented in Taylor–Hood finite elements; a Gaussian prior is derived from maximum-entropy constraints on velocity and pressure regularity, and additional soft constraints penalize the discretized continuity and momentum residuals. The MAP estimate is computed after analytical elimination of pressure, with no-slip walls enforced exactly and gradients obtained by finite-element assembly. Posterior uncertainty is quantified with a Gauss–Newton Laplace approximation and propagated to flow-rate, pressure-drop, and WSS quantities of interest. A weakly divergence-free variant (DFER) is also proposed. The method is evaluated on synthetic patient-specific cerebral and aortic geometries and compared with tricubic interpolation and a PINN baseline across SNRs and resolutions, with reported lower WSS errors in the region of interest and approximately 99% empirical coverage inside nominal 95% credible intervals.

Significance. The methodology is coherent and has clear practical advantages: it is training-free, avoids automatic differentiation and adjoint solves, uses a local FEM basis that is well suited to near-wall gradients, and provides analytic posterior uncertainty propagation for linear QoIs. The maximum-entropy prior construction and pressure elimination are internally consistent, and the comparison is broad across geometries, SNRs, and resolutions. However, the numerical evidence as presented is generated under an inverse-crime configuration: the truth is produced by the same Taylor–Hood discretization family and the same iid Gaussian noise model assumed by the reconstruction. The accuracy ranking and UQ calibration should therefore be viewed as a consistency check of an idealized pipeline, not yet as evidence for clinical applicability. With mismatched tests and a hyperparameter sensitivity analysis, the paper could be a strong contribution.

major comments (3)
  1. [Secs. 2.5 and 2.1; Appendix 5.1] The numerical evaluation is an inverse-crime setup. The ground truth is generated with a steady Navier–Stokes solver using Taylor–Hood (P2/P1) DOLFINx elements (Sec. 2.5), and the reconstruction space is the same Taylor–Hood pair (Sec. 2.1, Eqs. (2)–(3)); the noise is iid Gaussian with known σ0², exactly the likelihood in Sec. 2.1.2. The manuscript does not state whether the reconstruction mesh equals the CFD mesh, nor whether the observation operator T is pointwise or voxel-averaging. Under this setup the truth is nearly representable in the reconstruction space and lies in the support of the physics prior, so the reported errors, the ranking vs. baselines, and the ≈99% coverage in Secs. 3.2–3.3 are only established for zero model misspecification. The Sec. 4 limitation statement acknowledges synthetic data, but that is weaker than the actual issue: the synthetic data are generated with
  2. [Sec. 2.4] The hyperparameters µs=10⁻⁶, µc=200, µm=200, αv=αp=1 are fixed across all cases, and no sensitivity analysis is reported. The stated Morozov validation only checks that the final data misfit is within 5% of NOBSσ0²; it does not determine these weights or demonstrate that the reconstructions and uncertainty intervals are insensitive to them. Since the reported WSS errors and the ranking relative to baselines depend on the data/continuity/momentum balance, please provide a sensitivity study (e.g., vary µc and µm over one to two orders of magnitude, vary αv and αp, for at least one geometry at low and high SNR) or a systematic selection procedure.
  3. [Sec. 2.1.3, Eq. (15)] The Laplace covariance uses the Gauss–Newton Hessian, dropping the second derivative of the advection term to ensure positive semi-definiteness. At the inlet Reynolds numbers in Table 2.1 (up to ~1044), advection is not weak, and the claim that ξ is small because the momentum residual is penalized is not demonstrated. Since UQ calibration is a central claimed contribution, please quantify the effect of the dropped term for at least one representative case (e.g., compare full-Hessian and Gauss–Newton covariances, or use a small MCMC check) and report whether coverage changes materially.
minor comments (5)
  1. [Sec. 2.3.1, Eq. (18)] Eq. (18) as written enforces only the single global constraint ∫Ω ∇·v dΩ = 0, not the standard weak divergence-free condition ∫ q ∇·v dΩ = 0 for all q∈P^h. Please correct the displayed constraint or clarify what projection was actually used for the tricubic baseline.
  2. [Sec. 2.1.3, before Eq. (11)] The displayed log-posterior has the pressure smoothness term with a '+' sign; it should be '−' for the posterior to be proper. Eq. (11) itself is correct, but the intermediate expression should be fixed. Also, §2.1.1 introduces independent λ_v and λ_p, but Eq. (11) uses a single µ_s for both velocity and pressure smoothness blocks; please define λ_s or use separate weights.
  3. [Sec. 2.1] There are small typos: 'NVB=NV+NED and NVB=NV' should be 'NVB=NV+NED and NPB=NV'; the nondimensionalization line 'p=v/(ρV²)' should be 'p=p/(ρV²)'.
  4. [Sec. 2.4 and Sec. 3.3] The PINN loss weights are said to be chosen via Morozov's principle so that the data loss is O(1); this is not a standard use of Morozov's discrepancy principle. The timing comparison in Sec. 3.3 is informal and uses different hardware; please label it accordingly. Also state the number of noise realizations per setting in Sec. 2.5; if each setting is one sample, the small FER-vs-DFER differences in Appendix 5.1 may not be statistically meaningful.
  5. [Sec. 2.4] The statement that code will be available upon publication is not sufficient for reproducibility; please include the repository and ideally a runnable example for one case in the supplementary material.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation: MAP/UQ follow from stated max-entropy prior and Gaussian likelihood; self-citations are contextual; synthetic-data limitation is inverse-crime consistency, not equation-level circularity.

full rationale

The central derivation is self-contained. The prior in (10) is obtained by solving the max-entropy problems (5) and (9); the likelihood is the Gaussian noise model in Sec. 2.1.2; the posterior MAP is (11), pressure elimination (12), and Laplace covariance (14) follow by direct differentiation. No predicted QoI (velocity, pressure, WSS, flow rate) is a fitted parameter or an algebraic rearrangement of the data term. The regularization weights (mu_s, mu_c, mu_m) are fixed constants, and the Morozov check in Sec. 2.4 is an a posteriori validation rather than a fit to the QoIs. The benchmark comparison is external: truth is a CFD solution, baselines are tricubic interpolation and a PINN. The two self-citations ([33] Sautory & Shadden; [43] Wu, Wang & Shadden) are contextual -- one supports an intro claim about PINN training, the other offers an equivalent virtual-observation view of the already-derived prior -- and neither is load-bearing. The acknowledged synthetic-data/inverse-crime setup (Sec. 2.5 vs. Sec. 2.1) is a threat to external validity and is acknowledged in Sec. 4, but it does not make the derivation circular by construction.

Assumptions & free parameters 6 free parameters · 6 assumptions · 1 invented entities

The ledger shows the method's priors are parameterized by five hand-set scalars (µ_s, µ_c, µ_m, α_v, α_p) that enter the MAP objective and the Laplace covariance directly; their values are validated only by a discrepancy check, never by sensitivity analysis. The physics content is assumed (steady NS, known geometry, known σ0), and the test data are generated under those same assumptions. No new physical entities are introduced; the 'virtual observations' y_c, y_m are a bookkeeping device.

free parameters (6)
  • µ_s (smoothness weight) = 1e-6
    Hand-set for all experiments (Sec. 2.4); controls prior strength of the M+αK smoothness penalty via µ_s=σ0²λ_s/NOBS.
  • µ_c (continuity weight) = 200
    Hand-set; weights the projected divergence residual in the MAP objective (11).
  • µ_m (momentum weight) = 200
    Hand-set; weights the projected Navier-Stokes momentum residual.
  • α_v = 1
    Design parameter balancing L2 boundedness vs H1 smoothness in L_v (Sec. 2.1.1).
  • α_p = 1
    Same for the pressure form L_p.
  • Morozov discrepancy tolerance = 0.05
    Acceptance threshold for the fixed weights: |data misfit/(NOBS σ0²) - 1| ≤ 0.05 (Sec. 2.4); the check is post-hoc, not a parameter-selection loop.
assumptions (6)
  • domain assumption The steady incompressible Navier-Stokes equations (7) with a single known Reynolds number model the true blood flow.
    The physics prior penalizes residuals of (7) (Sec. 2.1.1); the Discussion admits the steady-flow restriction. If pulsatility, turbulence, or non-Newtonian rheology matter, the prior biases the reconstruction.
  • domain assumption Vessel geometry is known exactly; segmentation and no-slip wall location are given.
    No-slip is applied by zeroing wall DOFs and the domain Ωh is the given mesh (Sec. 2.1, Eq. 2); the Discussion lists known geometry as a limitation.
  • domain assumption Measurement noise is additive, iid Gaussian, with known variance σ0².
    Likelihood (Sec. 2.1.2) assumes Σ0=σ0²I with σ0 from machine specs; synthetic data are generated with exactly this model (Sec. 2.5), so mismatch with real PC-MRI noise is untested.
  • standard math The maximum-entropy/minimum-relative-entropy solutions (6) and (10) are the correct least-informative priors under the stated moment constraints.
    Standard convex information-theoretic derivation (Sec. 2.1.1, refs [41,42]); note the chosen quadratic forms L_v, L_p (M+αK) are design choices, and the strict-convexity claim that makes (δ_c,δ_m)→(λ_c,λ_m) one-to-one is asserted, not proven in detail.
  • domain assumption The quasi-Newton solver converges to the global maximizer of the posterior.
    The objective (11) is a quartic polynomial in the velocity DOFs (nonconvex through A([v̂])); convergence to a global MAP is asserted via the claim of a favorable optimization landscape (Sec. 2.1.3), without a guarantee.
  • domain assumption The Laplace/Gauss-Newton approximation captures posterior moments well enough for UQ.
    Requires posterior concentration near the MAP and small momentum residual; the higher-order Hessian term is dropped (Sec. 2.1.3). Empirically the intervals over-cover (~99% nominal 95%), consistent with conservatism.
invented entities (1)
  • Virtual constraint observations y_c, y_m
    purpose: Interpretive device that recasts the maximum-entropy physics prior as Bayesian conditioning on zero-valued virtual residuals with covariance (|Ω|/λ_c)M_p and (|Ω|/λ_m)M_v (Sec. 2.1.1).
    These are not physical entities; they are latent variables introduced to connect the prior to the authors' earlier constraint-update framework [43]. They carry no falsifiable predictions of their own and do not add physical content.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Bayesian finite element regression for vascular flow reconstruction with quantified uncertainty." pith.science (2026). https://pith.science/paper/AVPMOMM5

@misc{pith2026260720224,
  author       = {Pith},
  title        = {Pith review of: Bayesian finite element regression for vascular flow reconstruction with quantified uncertainty},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AVPMOMM5}},
  note         = {Machine review of arXiv:2607.20224}
}
read the original abstract

Reconstructing accurate velocity and pressure fields from under-resolved noisy measurements of blood flow is an ill-posed inverse problem due to unknown inlet and outlet boundary conditions. We present a Bayesian finite element regression framework that reconstructs steady three-dimensional velocity and pressure fields, with quantified uncertainty, from noisy velocity observations without offline training data. We represent velocity and pressure fields in Taylor-Hood finite element basis functions, and construct physics-informed priors on the nodal degrees of freedom from maximum-entropy principles. Combined with a likelihood specified by a noise-model, this yields a posterior whose maximum-a-posteriori estimate (MAP) gives velocity and pressure reconstructions. The MAP estimate is computed by solving a large-scale sparse nonlinear least-squares problem where pressure is eliminated analytically, no-slip walls are enforced exactly, and gradient is computed without forward/adjoint solves or automatic differentiation. A Laplace approximation of the posterior quantifies the uncertainties in our reconstructions and propagates them to clinically relevant quantities of interest including, pressure drop, flow rates, and wall shear stress. On patient-specific cerebral aneurysm, aortic aneurysm, and aortic coarctation geometries, the method reconstructs velocity and pressure more accurately than tricubic interpolation and comparably to a PINN, while recovering region-of-interest wall shear stress more accurately than both.

Figures

Figures reproduced from arXiv: 2607.20224 by the authors.

Figure 2.1
Figure 2.1. Geometries used in the numerical experiments. [PITH_FULL_IMAGE:figures/full_fig_p016_2_1.png] view at source ↗
Figure 3.1
Figure 3.1. Regions of interest for different geometries. [PITH_FULL_IMAGE:figures/full_fig_p017_3_1.png] view at source ↗
Figure 3.2
Figure 3.2. Observations, reconstructions, and corresponding error fields on a planar slice through the aortic aneurysm [PITH_FULL_IMAGE:figures/full_fig_p017_3_2.png] view at source ↗
Figures from the paper (8 more)
Figure 3.3
Figure 3.3. Figure 3.3: Observations, reconstructions, and corresponding error fields on a planar slice through the aortic aneurysm [PITH_FULL_IMAGE:figures/full_fig_p018_3_3.png]
Figure 3.4
Figure 3.4. Figure 3.4: Observations, reconstructions, and corresponding error fields on a planar slice through the cerebral aneurysm [PITH_FULL_IMAGE:figures/full_fig_p018_3_4.png]
Figure 3.5
Figure 3.5. Figure 3.5: Flow rates and average pressures, with corresponding uncertainty estimates, evaluated on planes orthogonal to [PITH_FULL_IMAGE:figures/full_fig_p019_3_5.png]
Figure 3.6
Figure 3.6. Figure 3.6: Reconstruction errors, posterior uncertainties, and coverage fields for the z-velocity, pressure, and WSS magnitude [PITH_FULL_IMAGE:figures/full_fig_p020_3_6.png]
Figure 3.7
Figure 3.7. Figure 3.7: Reconstruction errors, posterior uncertainties, and coverage fields for the z-velocity, pressure, and WSS magnitude [PITH_FULL_IMAGE:figures/full_fig_p021_3_7.png]
Figure 3.8
Figure 3.8. Figure 3.8: Reconstruction errors, posterior uncertainties, and coverage fields for the z-velocity, pressure, and WSS magnitude [PITH_FULL_IMAGE:figures/full_fig_p022_3_8.png]
Figure 3.9
Figure 3.9. Figure 3.9: Relative RMS reconstruction errors (%) for velocity, pressure, and WSS obtained with each method, averaged [PITH_FULL_IMAGE:figures/full_fig_p023_3_9.png]
Figure 3.10
Figure 3.10. Figure 3.10: Objective function value history for different geometries. Aortic aneurysm (left), aortic coarctation (middle), [PITH_FULL_IMAGE:figures/full_fig_p024_3_10.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

51 extracted references

  1. [1]

    4d flow cardiovascular magnetic resonance consensus statement: 2023 update,

    M. M. Bissell, F. Raimondi, L. Ait Ali, B. D. Allen, A. J. Barker, A. Bolger, N. Burris, C.-J. Carhäll, J. D. Collins, T. Ebbers, C. J. Francois, A. Frydrychowicz, P. Garg, J. Geiger, H. Ha, A. Hennemuth, M. D. Hope, A. Hsiao, K. Johnson, S. Kozerke, L. E. Ma, M. Markl, D. Martins, M. Messina, T. H. Oechtering, P. van Ooij, C. Rigsby, J. Rodriguez-Palomar...

    2023

  2. [2]

    4d flow mri,

    M. Markl, A. Frydrychowicz, S. Kozerke, M. Hope, and O. Wieben, “4d flow mri,”Journal of Magnetic Resonance Imaging, vol. 36, no. 5, pp. 1015–1036, 2012

    2012

  3. [3]

    Comparison of 4d flow and 2d velocity-encoded phase contrast mri sequences for the evaluation of aortic hemodynamics,

    E. Bollache, P. van Ooij, A. Powell, J. Carr, M. Markl, and A. J. Barker, “Comparison of 4d flow and 2d velocity-encoded phase contrast mri sequences for the evaluation of aortic hemodynamics,” International Journal of Cardiovascular Imaging, vol. 32, no. 10, pp. 1529–1541, 2016

    2016

  4. [4]

    Ultrasound vector flow imaging-part i: Sequential systems,

    J. A. Jensen, S. I. Nikolov, A. C. H. Yu, and D. Garcia, “Ultrasound vector flow imaging-part i: Sequential systems,”IEEE transactions on ultrasonics, ferroelectrics, and frequency control, vol. 63, no. 11, pp. 1704–1721, 2016

    2016

  5. [5]

    Ultrasound vector flow imaging-part ii: Parallel systems,

    J. A. Jensen, S. I. Nikolov, A. C. H. Yu, and D. Garcia, “Ultrasound vector flow imaging-part ii: Parallel systems,”IEEE transactions on ultrasonics, ferroelectrics, and frequency control, vol. 63, no. 11, pp. 1722–1732, 2016

    2016

  6. [6]

    Unifying perspective for gappy proper orthogonal decomposition and probabilistic principal component analysis,

    K. Lee and D. N. Mavris, “Unifying perspective for gappy proper orthogonal decomposition and probabilistic principal component analysis,”AIAA journal, vol. 48, no. 6, pp. 1117–1129, 2010. 25

    2010

  7. [7]

    Proper orthogonal decomposition truncation method for data denoising and order reduction,

    M. C. Brindise and P. P. Vlachos, “Proper orthogonal decomposition truncation method for data denoising and order reduction,”Experiments in fluids, vol. 58, no. 4, pp. 1–18, 2017

    2017

  8. [8]

    Denoising and spatial resolution enhancement of 4d flow mri using proper orthogonal decomposition and lasso regularization,

    M. F. Fathi, A. Bakhshinejad, A. Baghaie, D. Saloner, R. H. Sacho, V. L. Rayz, and R. M. D’Souza, “Denoising and spatial resolution enhancement of 4d flow mri using proper orthogonal decomposition and lasso regularization,”Computerized medical imaging and graphics, vol. 70, pp. 165–172, 2018

    2018

Show all 51 references
  1. [9]

    Towards reduced order models via robust proper orthogonal decomposition to capture personalised aortic haemody- namics,

    C. Chatpattanasiri, G. Franzetti, M. Bonfanti, V. Diaz-Zuccarini, and S. Balabani, “Towards reduced order models via robust proper orthogonal decomposition to capture personalised aortic haemody- namics,”Journal of biomechanics, vol. 158, pp. 111759–111759, 2023

    2023

  2. [10]

    Frequency–time analysis, low-rank reconstruction and denoising of turbulent flows using spod,

    A. Nekkanti and O. T. Schmidt, “Frequency–time analysis, low-rank reconstruction and denoising of turbulent flows using spod,”Journal of fluid mechanics, vol. 926, 2021

    2021

  3. [11]

    Data-driven cardiovascular flow modelling: examples and opportu- nities,

    A. Arzani and S. T. M. Dawson, “Data-driven cardiovascular flow modelling: examples and opportu- nities,”Journal of the Royal Society Interface, vol. 18, no. 175, pp. 20200802–, 2021

    2021

  4. [12]

    Dynamicmodedecomposition as a framework for denoising ultrafast power doppler images,

    B.Pialot, F.Guidi, P.Muleki-Seya, E.Boni, A.Ramalli, andF.Varray, “Dynamicmodedecomposition as a framework for denoising ultrafast power doppler images,”Computer methods and programs in biomedicine, vol. 276, pp. 109221–, 2026

    2026

  5. [13]

    Tricubic interpolation in three dimensions,

    F. Lekien and J. Marsden, “Tricubic interpolation in three dimensions,”International journal for numerical methods in engineering, vol. 63, no. 3, pp. 455–471, 2005

    2005

  6. [14]

    Reconstruction of divergence-free velocity fields from cine 3d phase-contrast flow measurements,

    J. Busch, D. Giese, L. Wissmann, and S. Kozerke, “Reconstruction of divergence-free velocity fields from cine 3d phase-contrast flow measurements,”Magnetic Resonance in Medicine, vol. 69, no. 1, pp. 200–210, 2013

    2013

  7. [15]

    Robust 4d flow denoising using divergence-free wavelet transform,

    F. Ong, M. Uecker, U. Tariq, A. Hsiao, M. T. Alley, S. S. Vasanawala, and M. Lustig, “Robust 4d flow denoising using divergence-free wavelet transform,”Magnetic Resonance in Medicine, vol. 73, no. 2, pp. 828–842, 2015

    2015

  8. [16]

    Simvascular: An open source pipeline for cardiovascular simulation,

    A. Updegrove, N. M. Wilson, J. Merkow, H. Lan, A. L. Marsden, and S. C. Shadden, “Simvascular: An open source pipeline for cardiovascular simulation,”Annals of biomedical engineering, vol. 45, no. 3, pp. 525–541, 2017

    2017

  9. [17]

    Crimson: An open-source software framework for cardiovascular integrated modelling and simulation,

    C. J. Arthurs, R. Khlebnikov, A. Melville, M. Marčan, A. Gomez, D. Dillon-Murphy, F. Cuomo, M. Silva Vieira, J. Schollenberger, S. R. Lynch, C. Tossas-Betancourt, K. Iyer, S. Hopper, E. Liv- ingston, P. Youssefi, A. Noorani, S. Ben Ahmed, F. J. H. Nauta, T. M. J. van Bakel, Y....

    2021

  10. [18]

    Boundary control in computational haemodynamics,

    T. S. Koltukluoğlu and P. J. Blanco, “Boundary control in computational haemodynamics,”Journal of Fluid Mechanics, vol. 847, pp. 329–364, 2018

    2018

  11. [19]

    Fourier spectral dynamic data assimilation: Interlacing cfd with 4d flow mri,

    T. S. Koltukluoğlu, T. Liu, T. M. Peters, P.-T. Yap, C. Essert, A. Khan, D. Shen, L. H. Staib, and S. Zhou, “Fourier spectral dynamic data assimilation: Interlacing cfd with 4d flow mri,” inMedical Image Computing and Computer Assisted Intervention - MICCAI 2019, vol. 11765 of...

    2019

  12. [20]

    Harmonic balance techniques in cardiovascular fluid mechanics,

    T. S. Koltukluoğlu, G. Cvijetić, R. Hiptmair, T. Liu, T. M. Peters, P.-T. Yap, C. Essert, A. Khan, D. Shen, L. H. Staib, and S. Zhou, “Harmonic balance techniques in cardiovascular fluid mechanics,” in Medical Image Computing and Computer Assisted Intervention - MICCAI 2019, v...

    2019

  13. [21]

    Variational data assimilation for transient blood flow simulations: Cerebral aneurysms as an illustrative example,

    S. W. Funke, M. Nordaas, O. Evju, M. S. Alnaes, and K. A. Mardal, “Variational data assimilation for transient blood flow simulations: Cerebral aneurysms as an illustrative example,”International Journal for Numerical Methods in Biomedical Engineering, vol. 35, no. 1, pp. e315...

    2019

  14. [22]

    Joint reconstruction and segmentationofnoisyvelocityimagesasaninversenavier–stokesproblem,

    A. Kontogiannis, S. V. Elgersma, A. J. Sederman, and M. P. Juniper, “Joint reconstruction and segmentationofnoisyvelocityimagesasaninversenavier–stokesproblem,”Journal of fluid mechanics, vol. 944, 2022

    2022

  15. [23]

    Bayesian inverse navier–stokes problems: joint flow field reconstruction and parameter learning,

    A. Kontogiannis, S. V. Elgersma, A. J. Sederman, and M. P. Juniper, “Bayesian inverse navier–stokes problems: joint flow field reconstruction and parameter learning,”Inverse problems, vol. 41, no. 1, pp. 15008–, 2025

    2025

  16. [24]

    Enhancement of cerebrovascular 4d flow mri velocity fields using machine learning and computational fluid dynamics simulation data,

    D. R. Rutkowski, A. Roldán-Alzate, and K. M. Johnson, “Enhancement of cerebrovascular 4d flow mri velocity fields using machine learning and computational fluid dynamics simulation data,”Scientific Reports, vol. 11, no. 1, pp. 10240–11, 2021

    2021

  17. [25]

    Srflow: Deep learning based super-resolution of 4d-flow mri data,

    S. Shit, J. Zimmermann, I. Ezhov, J. C. Paetzold, A. F. Sanches, C. Pirkl, and B. H. Menze, “Srflow: Deep learning based super-resolution of 4d-flow mri data,”Frontiers in artificial intelligence, vol. 5, pp. 928181–, 2022

    2022

  18. [26]

    Mesh-based super-resolution of fluid flows with multiscale graph neural networks,

    S. Barwey, P. Pal, S. Patel, R. Balin, B. Lusch, V. Vishwanath, R. Maulik, and R. Balakrishnan, “Mesh-based super-resolution of fluid flows with multiscale graph neural networks,”Computer methods in applied mechanics and engineering, vol. 443, pp. 118072–, 2025

    2025

  19. [27]

    Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations,

    M. Raissi, P. Perdikaris, and G. Karniadakis, “Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations,” Journal of Computational Physics, vol. 378, pp. 686–707, 2019

    2019

  20. [28]

    Physics-informed machine learning,

    G. E. Karniadakis, I. G. Kevrekidis, L. Lu, P. Perdikaris, S. Wang, and L. Yang, “Physics-informed machine learning,”Nature reviews physics, vol. 3, no. 6, pp. 422–440, 2021

    2021

  21. [29]

    Super- resolution and denoising of 4d-flow mri using physics-informed deep neural nets,

    M. F. Fathi, I. Perez-Raya, A. Baghaie, P. Berg, G. Janiga, A. Arzani, and R. M. D’Souza, “Super- resolution and denoising of 4d-flow mri using physics-informed deep neural nets,”Computer Methods and Programs in Biomedicine, vol. 197, pp. 105729–, 2020

    2020

  22. [30]

    Physics-constrained bayesian neural network for fluid flow reconstruction with sparse and noisy data,

    L. Sun and J.-X. Wang, “Physics-constrained bayesian neural network for fluid flow reconstruction with sparse and noisy data,”Theoretical and applied mechanics letters, vol. 10, no. 3, pp. 161–169, 2020

    2020

  23. [31]

    Super-resolution and denoising of fluid flow using physics-informed convolutional neural networks without high-resolution labels,

    H. Gao, L. Sun, and J.-X. Wang, “Super-resolution and denoising of fluid flow using physics-informed convolutional neural networks without high-resolution labels,”Physics of Fluids, vol. 33, no. 7, pp. 73603–, 2021

    2021

  24. [32]

    Physics-informed graph neural galerkin networks: A unified framework for solving pde-governed forward and inverse problems,

    H. Gao, M. J. Zahr, and J.-X. Wang, “Physics-informed graph neural galerkin networks: A unified framework for solving pde-governed forward and inverse problems,”Computer methods in applied mechanics and engineering, vol. 390, pp. 114502–, 2022

    2022

  25. [33]

    Unsupervised denoising and super-resolution of vascular flow data by physics-informed machine learning,

    T. Sautory and S. C. Shadden, “Unsupervised denoising and super-resolution of vascular flow data by physics-informed machine learning,”Journal of biomechanical engineering, vol. 146, no. 9, 2024. 27

    2024

  26. [34]

    Flow reconstruction by multiresolution optimization of a discrete loss with automatic differentiation,

    P. Karnakov, S. Litvinov, and P. Koumoutsakos, “Flow reconstruction by multiresolution optimization of a discrete loss with automatic differentiation,”The European physical journal. E, Soft matter and biological physics, vol. 46, no. 7, pp. 59–, 2023

    2023

  27. [35]

    Solving inverse problems in physics by optimizing a discrete loss: Fast and accurate learning without neural networks,

    P. Karnakov, S. Litvinov, and P. Koumoutsakos, “Solving inverse problems in physics by optimizing a discrete loss: Fast and accurate learning without neural networks,”PNAS nexus, vol. 3, no. 1, pp. pgae005–, 2024

    2024

  28. [36]

    Bayesian inference for pde-based inverse problems using the optimization of a discrete loss,

    L. Amoudruz, S. Litvinov, C. Papadimitriou, and P. Koumoutsakos, “Bayesian inference for pde-based inverse problems using the optimization of a discrete loss,”Computer methods in applied mechanics and engineering, vol. 455, pp. 118903–, 2026

    2026

  29. [37]

    P. B. Bochev and M. D. Gunzburger,Least-squares finite element methods / Pavel B. Bochev, Max D. Gunzburger.Applied mathematical sciences ; 166, New York: Springer, 2009

    2009

  30. [38]

    Dataassimilationfornavier-stokesusingtheleast-squaresfinite-element method,

    A.SchwarzandR.P.Dwight, “Dataassimilationfornavier-stokesusingtheleast-squaresfinite-element method,”International journal for uncertainty quantification, vol. 8, no. 5, pp. 383–403, 2018

    2018

  31. [39]

    Weighted least- squares finite elements based on particle imaging velocimetry data,

    J. Heys, T. Manteuffel, S. McCormick, M. Milano, J. Westerdale, and M. Belohlavek, “Weighted least- squares finite elements based on particle imaging velocimetry data,”Journal of computational physics, vol. 229, no. 1, pp. 107–118, 2010

    2010

  32. [40]

    Bayesian inference for data assimilation using least-squares finite element methods,

    R. P. Dwight, “Bayesian inference for data assimilation using least-squares finite element methods,” IOP conference series. Materials Science and Engineering, vol. 10, no. 1, pp. 012224–012224, 2010

    2010

  33. [41]

    Skilling,Maximum Entropy and Bayesian Methods : Cambridge, England, 1988 / edited by John Skilling.Fundamental Theories of Physics, 36, Dordrecht: Springer Netherlands, 1st ed

    J. Skilling,Maximum Entropy and Bayesian Methods : Cambridge, England, 1988 / edited by John Skilling.Fundamental Theories of Physics, 36, Dordrecht: Springer Netherlands, 1st ed. 1989. ed., 1989

    1988

  34. [42]

    Entropy, information, and the updating of probabilities,

    A. Caticha, “Entropy, information, and the updating of probabilities,”Entropy (Basel, Switzerland), vol. 23, no. 7, pp. 895–, 2021

    2021

  35. [43]

    Adding constraints to bayesian inverse problems,

    J. Wu, J.-X. Wang, and S. C. Shadden, “Adding constraints to bayesian inverse problems,” inPro- ceedings of the ... AAAI Conference on Artificial Intelligence, vol. 33, pp. 1666–1673, 2019

    2019

  36. [44]

    Supplementary material. dolfinx: The next generation fenics problem solving environment,

    I. Baratta, J. Dean, J. S. Dokken, M. Habera, J. S. Hale, C. Richardson, M. E. Rognes, M. W. Scroggs, N. Sime, and G. N. Wells, “Supplementary material. dolfinx: The next generation fenics problem solving environment,” 2023

    2023

  37. [45]

    Construction of arbitrary order finite element degree-of-freedom maps on polygonal and polyhedral cell meshes,

    M. W. Scroggs, J. S. Dokken, C. N. Richardson, and G. N. Wells, “Construction of arbitrary order finite element degree-of-freedom maps on polygonal and polyhedral cell meshes,”ACM Transactions on Mathematical Software, vol. 48, no. 2, pp. 1–23, 2022

    2022

  38. [46]

    Basix: a runtime finite element basis evaluation library,

    M. W. Scroggs, I. A. Baratta, C. N. Richardson, and G. N. Wells, “Basix: a runtime finite element basis evaluation library,”Journal of open source software, vol. 7, no. 73, pp. 3982–, 2022

    2022

  39. [47]

    Unified form language: A domain-specific language for weak formulations of partial differential equations,

    M. S. Alnæs, A. Logg, K. B. Ølgaard, M. E. Rognes, and G. N. Wells, “Unified form language: A domain-specific language for weak formulations of partial differential equations,”ACM transactions on mathematical software, vol. 40, no. 2, pp. 1–37, 2014

    2014

  40. [48]

    Parallel distributed computing using python,

    L. D. Dalcin, R. R. Paz, P. A. Kler, and A. Cosimo, “Parallel distributed computing using python,” Advances in water resources, vol. 34, no. 9, pp. 1124–1139, 2011. 28

    2011

  41. [49]

    Balay, S

    S. Balay, S. Abhyankar, M. F. Adams, J. Brown, P. Brune, K. Buschelman, E. M. Constantinescu, L. Dalcin, S. Benson, A. Dener, V. Eijkhout, J. Faibussowitsch, W. D. Gropp, V. Hapla, T. Isaac, P. Jolivet, D. Karpeev, D. Kaushik, M. G. Knepley, F. Kong, S. Kruger, D. A. May, L. C...

    2025

  42. [50]

    Arbtools: A tricubic spline interpolator for three-dimensional scalar or vector fields,

    P. Walker, U. Krohn, and D. Carty, “Arbtools: A tricubic spline interpolator for three-dimensional scalar or vector fields,”Journal of open research software, vol. 7, no. 1, pp. 12–, 2019. 29

    2019

  43. [51]

    Performance tables for the proposed approach and the baseline methods

    Appendix 5.1. Performance tables for the proposed approach and the baseline methods. The tables below summarize the relative (%) reconstruction errors in velocity, pressure, and wall shear stress (WSS). For each (SNR, resolution) pair the top cell contains the global error ove...

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.