Pith. sign in

REVIEW 6 major objections 5 minor 48 references

Motion-compensated cardiac MRI using low-rank diffeomorphic flow (DMoCo)

T0 review · 6 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read A low-rank diffeomorphic flow model, learned directly from undersampled k-space data, reconstructs free-breathing, ungated 3D cardiac MRI into cardiac- and respiratory-resolved cine from a six-minute scan.

desk verdict Novel low-rank diffeomorphic flow model for 5D cardiac MRI, with convincing phantom evidence but an unaddressed contrast-drift confound that weakens the in-vivo claim. read the letter →

arxiv 2505.03149 v3 pith:KGAK7M2B submitted 2025-05-06 cs.CV cs.AI

Joseph Kettelkamp , Ludovica Romanin , Sarv Priya , Mathews Jacob This is my paper · ORCID
classification cs.CVcs.AI
keywords cardiacMRImotioncompensationdiffeomorphicregistrationlow-rankmodelfree-breathingself-supervisedreconstructionradialk-space5Dimaging
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

This paper claims that six minutes of free-breathing, ungated 3D cardiac MRI can be turned into a cardiac- and respiratory-resolved 3D movie without breath holds, ECG gating, or supervised training data. The key move is to represent every phase volume as a fixed static template deformed by a diffeomorphism, with the deformations obtained by integrating a phase-parametrized velocity field that is low-rank across phases. The template and motion parameters are learned directly from undersampled k-space by minimizing data consistency plus penalties for flow path independence and template smoothness. On a simulated phantom the method recovers diaphragm displacement as 26.06 millimeters against a 27.33-millimeter ground truth, and it reports smoother time profiles and better respiratory estimates than the comparison methods; on five patients the functional measures track 2D cine. If correct, this would make cardiac functional MRI available to patients who cannot hold their breath and would deliver deformation maps usable for strain analysis.

What carries the argument

The load-bearing object is the low-rank velocity tensor $v_\tau(s,r)=P(r)M_\kappa(s,\tau)$, in which the spatial basis $P$ is shared by all phases and the phase-dependent weights are produced by a small multilayer perceptron. The deformation at phase $\tau$ is the endpoint of the ODE $d\phi_\delta/d\delta=\langle v_{\gamma(\delta)}(\phi_\delta),\,\gamma'(\delta)\rangle$ integrated along a straight-line path $\gamma$ from the template phase to $\tau$, and an alternate randomized path $\beta$ supplies the path-independence penalty $\mathbb{E}_{D_\tau}\|\psi_\tau-\varphi_\tau\|^2$. Because every deformation is built from the same velocity basis through overlapping integrals, the model couples all phases together and is therefore more constrained than a direct low-rank deformation model.

What would settle it

Simulate the same free-breathing radial acquisition on a numerical phantom with known motion, then add a slow blood-pool intensity drift (for instance a ten percent linear gain over six minutes) and check whether the recovered deformations and template shift to compensate; if they do, the model has confused contrast change with motion.

Watch

Extended reading notes

Core claim

The central claim is that tissue velocities, rather than images or deformations, are the right low-rank object for jointly modeling cardiac and respiratory motion. The paper writes the velocity tensor at phase $\tau$ as $v_\tau(s,r)=\sum_{n=0}^R p_n(r)m_n(s,\tau)$, represents the image at phase $\tau$ as $\rho_t(r)=\eta(\varphi_{\tau(t)}(r))$, and computes each deformation $\varphi_\tau$ by integrating the velocity field along a straight path through the two-dimensional phase space from the template to $\tau$. A path-independence penalty compares that deformation with the one obtained along a randomly perturbed path, which further constrains the family. The static template $\eta$ and the spatial basis $P$ and MLP weights $\kappa$ are all fitted from undersampled k-space by stochastic gradient descent, with no fully sampled reference. The paper argues that overlapping path integrals and the path penalty make this deformation family smoother and more constrained than directly parameterized deformations, and that is why its reconstructed time profiles are less jumpy and its respiratory displacement estimates are closer to ground truth.

Load-bearing premise

The method assumes every measured image is exactly one fixed static template seen through a smooth invertible deformation, so any time-varying intensity change—such as contrast washout over the six-minute scan—has no separate representation and would be silently absorbed into false motion.

Editorial extensions

If this is right

  • Free-breathing, ungated acquisitions of roughly six minutes can yield cardiac- and respiratory-resolved 3D cine, removing the breath hold and ECG gating requirements of current cardiac functional MRI.
  • Because the velocity tensor needs a much lower rank than images or deformation fields, motion estimates should stay stable under heavy undersampling and low contrast-to-noise ratio.
  • The path-independence penalty suppresses oscillatory, nonphysical deformations and abrupt jumps in the temporal profiles, which is what the paper says separates it from direct deformation models.
  • The reconstruction directly outputs smooth deformation fields for each cardiac and respiratory phase, which are available for myocardial strain analysis even though that analysis is future work.

Reading between the lines

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

  • Going beyond the paper, a direct test of the contrast-drift risk would be to simulate the same six-minute radial acquisition with a slow blood-pool intensity drift and check whether the recovered motion fields absorb the drift; the paper's model has no mechanism to distinguish intensity change from deformation.
  • The same phase-space flow machinery should extend to other quasi-periodic motions in MRI, such as abdominal compression or fetal motion, whenever a low-dimensional phase coordinate can be measured from navigators or self-gating.
  • The path-independence penalty may erase genuine route dependence if cardiac contraction exhibits hysteresis; a phantom with known path-dependent motion could reveal whether this regularization is too strong.
  • The roughly three-hour reconstruction time on one GPU is the main practical obstacle the paper concedes; faster optimization or learned initialization would be needed before routine clinical use.
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

6 major / 5 minor

Summary. The paper introduces DMoCo, an unsupervised motion-compensated reconstruction algorithm for free-breathing, ungated 3D cardiac MRI. Each phase volume is modeled as a deformation of a static template, where the deformation is obtained by integrating a velocity field along a path in a two-dimensional phase space (cardiac and respiratory). The velocity field is represented with a low-rank model combining spatial basis functions and an MLP that maps phase variables to weights; a path-independence penalty further constrains the diffeomorphisms. The template and motion parameters are learned directly from k-space data via a data-consistency loss. Experiments on an XCAT phantom and a 2D annulus phantom show that low-rank velocity representations are more compact than low-rank deformation or image representations, and comparisons against a reimplemented MoCo-SToRM baseline and a motion-resolved method suggest improved recovery. In-vivo results on five patients provide qualitative comparisons with 2D cine and preliminary left-ventricular functional measures.

Significance. If the claims hold, the paper contributes a principled and compact motion model that could improve motion-compensated cardiac MRI in low-CNR, heavily undersampled settings. The mathematical formulation is coherent, the unsupervised formulation avoids the need for supervised training data, and the path-independence penalty is a novel regularization. The phantom experiments provide quantitative support for the core modeling claim, and the use of an XCAT phantom with realistic acquisition settings is a strength. However, the validation is at the feasibility level: the phantom comparisons use a single noise realization, the in-vivo evaluation is largely qualitative, and the comparison baseline is a reimplemented variant of MoCo-SToRM rather than the original method. The paper therefore offers a promising algorithmic idea whose practical advantage is not yet fully established.

major comments (6)
  1. [Sec. III-A (Eq. 5) and Sec. V-C] The model assumes each phase volume is exactly a deformation of a static template with no phase-dependent intensity or contrast variation. The in-vivo data were acquired 15–20 minutes after gadolinium injection and over a fixed 5:58 minute scan, so blood-pool and myocardial contrast can drift within the acquisition. Under the deformation-only model of Eq. (5), any such contrast drift would be absorbed into spurious deformations, biasing the template, motion fields, respiratory displacement estimates, and the functional measures in Fig. 7. The limitation paragraph (Sec. VII) mentions low CNR, fat suppression, and runtime but does not address this intensity-versus-deformation identifiability problem. The authors should either justify that contrast is effectively stationary over the acquisition window or demonstrate with a phantom experiment that includes time-varying intensity that the model does not introduce significant bias.
  2. [Sec. VI-C and Fig. 4] The quantitative XCAT comparison appears to be based on a single noise realization; no error bars or repeated trials are reported for the PSNR, SSIM, or diaphragm displacement values (e.g., 26.06 mm vs. 27.33 mm ground truth for DMoCo, 21.69 mm for MoCo-SToRM). Some differences, such as PSNR values of 29.87 dB vs. 29.69 dB for a given view, are small relative to typical run-to-run variability in stochastic optimization. To support the claim of improved recovery, the authors should report mean and standard deviation over multiple noise realizations and, if feasible, a significance test.
  3. [Sec. IV-E and Eq. (14)] The MoCo-SToRM baseline is reimplemented with a low-rank deformation model (Eq. 14) rather than using the original CNN-based MoCo-SToRM from [23]. This changes the baseline architecture and optimization landscape; the observed improvement over this reimplementation may reflect differences in network capacity, initialization, or hyperparameter tuning rather than the proposed low-rank velocity representation. The authors should compare against the original MoCo-SToRM implementation, or explicitly justify that the reimplementation is a faithful and competitive proxy.
  4. [Sec. IV-B and Eq. (11)] The loss in Eq. (11) involves several hyperparameters (lambda1, lambda2, lambda3, rank R, K=20, path-perturbation variance), and the paper states these were chosen by trial and error on one dataset and kept fixed for other datasets. No sensitivity analysis is provided in the main text or appendices. Since the central claim is that the additional constraints (path independence and low-rank velocity) improve results, the paper would benefit from showing that the performance gain over MoCo-SToRM is robust to reasonable variations in these parameters, particularly lambda1 and R.
  5. [Sec. VI-A and Fig. 2] The low-rank comparison in Fig. 2 uses velocity tensors estimated from the reference images via the proposed model (Eq. 13). Thus, the finding that velocities are more low-rank than deformations or images may be partly a consequence of the parameterization used to estimate them, rather than an intrinsic property of the underlying motion. The authors should either use analytically known velocities from the annulus phantom (where the motion is generated by known circle contractions) or clarify that the comparison is between fitted representations, which weakens the suggestion that the low-rank velocity model is inherently more compact.
  6. [Sec. VI-D and Fig. 7] The in-vivo validation against 2D cine uses only five patients and reports functional measures without statistical quantification (e.g., correlation coefficients, Bland-Altman limits, or confidence intervals). The plots show a best-fit line through the origin but no measures of agreement or variability. Given the abstract claims improved recovery over current motion-resolved and motion-compensated algorithms, the in-vivo comparison should either include a quantitative comparison against the SOTA baselines or be scoped more modestly as a feasibility demonstration with qualitative support.
minor comments (5)
  1. [Sec. II-B] The sentence 'This intuition has led to the representation of diffeomorphisms, which are often expressed as the endpoint of a flow' is awkwardly phrased and should be rewritten for clarity.
  2. [Eqs. (2)-(3)] The notation phi^1(r) and phi^0(r) is used without explicit definition; the superscripts denote the integration endpoint and should be explained.
  3. [Sec. IV-A, Eq. (10)] The auto-encoder loss in Eq. (10) appears to have a subscript inconsistency: the decoder is written as F_{theta1} while the encoder is E_{theta2}, which is swapped relative to the text. Please check and correct the notation, and ensure the loss is fully specified.
  4. [Sec. VI-C] The phrase 'piecewise-smooth profile with abrupt jumps' is contradictory; it should be 'piecewise-constant with abrupt jumps' or 'smooth with abrupt transitions'.
  5. [Fig. 3 caption] The formatting of PSNR values, e.g., '29.60±(0.38)', is unusual and inconsistent; the parenthetical standard deviation notation should be clarified or made uniform.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the reconstruction is fitted to measured k-space data and scored against external XCAT ground truth and 2D cine references.

full rationale

The claimed derivation chain is not circular. The template η and the low-rank velocity parameters P, κ are explicitly fitted to the measured k-space data through the data-consistency term in Eq. (11), so the abstract's statement that they are 'learned directly from the k-space data' accurately describes the fitting objective rather than disguising a fit as a prediction. The main evaluation claims are anchored to external references: the XCAT phantom has known ground-truth volumes, and the reported diaphragm displacement (26.06 mm versus 27.33 mm ground truth) is measured against that external truth; the in-vivo functional measures are compared with breath-held 2D cine, an independent reference protocol. The low-rank velocity claim is supported by the descriptive rank-versus-error experiment in Fig. 2, which motivates the model but does not by construction produce the reconstruction results. The self-citations to MoCo-SToRM [23] and the authors' MICCAI 2023 5D work [26] appear as baseline methods and as acquisition or sequence references; no load-bearing uniqueness theorem is imported from these works, and the baselines are evaluated against the same external metrics. Equation (5) is a modeling assumption that excludes phase-dependent intensity or contrast changes; the Section VII limitations discuss low CNR and fat suppression but do not test contrast-drift identifiability. That is a correctness or robustness limitation, not a circular reduction of the output to the input. No equation reproduces a target result by construction, and no fitted parameter is renamed as a prediction.

Assumptions & free parameters 6 free parameters · 6 assumptions · 0 invented entities

The central claim rests on (a) the ODE-flow construction of diffeomorphisms, (b) the single-template pure-deformation assumption, (c) linear superposition of cardiac and respiratory velocities, (d) the low-rank and spatial smoothness of the velocity family, (e) the path-independence prior, and (f) the forward NUFFT model. The main hand-set free parameters are the regularization weights lambda1, lambda2, lambda3, the model rank R, the readout group size K, and the self-gating pre-processing choices. The paper introduces no new physical entities; the velocity tensor is a standard re-parameterization. The least supported premise is the low-rank property in realistic 3D cardiac motion, tested only on a 2D annulus, and the absence of intensity variation in the template model.

free parameters (6)
  • lambda1 (path-independence penalty weight)
    Weight on the velocity-constraint term in Eq. (11); chosen by trial and error on one dataset (Sec. IV-B) and only constrained to be 'large enough'.
  • lambda2 (template smoothness weight)
    Weight on the template smoothness penalty in Eq. (11); tuned by trial and error to balance streak artifacts against blurring (Sec. IV-B).
  • lambda3 (initialization smoothness weight)
    Regularizer in the binned initialization reconstruction (Eq. 12), set to a high value to obtain blurred alias-free images; its value is not reported.
  • Rank R of the velocity model
    Rank in Eq. (9) controls the capacity and smoothness of the motion model; it is reported as 10 only for Fig. 3, and the ranks used for the XCAT and in-vivo comparisons are not given.
  • Number of readout groups K per iteration = 20
    K=20 readout groups closest to each sampled phase tau define the stochastic gradient step (Sec. IV-B); hand-set.
  • Self-gating and path-perturbation choices (1 Hz filter, auto-encoder with 512/256 hidden units, 10 percent path noise)
    Hand-set components of the self-gating stage (Sec. IV-A) and of the stochastic path perturbation in the velocity constraint (Sec. IV-B); their accuracy or sensitivity is not quantified.
assumptions (6)
  • standard math Diffeomorphic flows solve the ODE system (Eqs. 2-3), so integrating a smooth velocity field yields an invertible, smooth deformation.
    Background from Christensen et al. [28] and Beg et al. [29], invoked in Sec. II-B and used throughout to construct phi_tau and to justify invertibility.
  • domain assumption The tissue velocity at a phase is the linear superposition of the cardiac and respiratory velocity contributions (Sec. III-B: 'we assume linearity').
    Cardiac contraction and respiratory displacement are assumed to sum linearly in the velocity tensor; nonlinear coupling between the two motions would invalidate Eq. (6).
  • domain assumption Every phase volume is a pure geometric deformation of a single static template with no phase-dependent intensity or contrast change (Eq. 5).
    The entire recovery is anchored to Eq. (5); contrast drift or inflow changes would be encoded as spurious deformation. The paper never tests this, and the data were acquired after gadolinium injection over roughly six minutes.
  • domain assumption The family of velocity tensors across phases is well approximated by a low-rank model with smooth spatial basis on a 32^3 grid (Eq. 9).
    Justified only by the 2D annulus experiment in Fig. 2; no derivation and no 3D in-vivo verification of the low-rank property is provided, and real cardiac motion includes torsion and through-plane components.
  • ad hoc to paper Deformations should be independent of the integration path, enforced by the soft penalty term in Eq. (11).
    The path-independence property is a soft regularization choice, not derived from physiology; it reduces degrees of freedom but biases the estimate if the true deformation depends on the path.
  • domain assumption The multichannel k-space forward model (Eq. 1) with NUFFT and coil sensitivities describes the acquisition, and coil sensitivities are known or estimated reliably.
    Data consistency in Eq. (11) relies on the forward model in Eq. (1); the paper does not state how coil sensitivities are estimated or whether they are updated during optimization.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Motion-compensated cardiac MRI using low-rank diffeomorphic flow (DMoCo)." pith.science (2026). https://pith.science/paper/KGAK7M2B

@misc{pith2026250503149,
  author       = {Pith},
  title        = {Pith review of: Motion-compensated cardiac MRI using low-rank diffeomorphic flow (DMoCo)},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KGAK7M2B}},
  note         = {Machine review of arXiv:2505.03149}
}
read the original abstract

We introduce an unsupervised motion-compensated image reconstruction algorithm for free-breathing and ungated 3D cardiac magnetic resonance imaging (MRI). We express the image volume corresponding to each specific motion phase as the deformation of a single static image template. The main contribution of the work is the low-rank model for the compact joint representation of the family of diffeomorphisms, parameterized by the motion phases. The diffeomorphism at a specific motion phase is obtained by integrating a parametric velocity field along a path connecting the reference template phase to the motion phase. The velocity field at different phases is represented using a low-rank model. The static template and the low-rank motion model parameters are learned directly from the k-space data in an unsupervised fashion. The more constrained motion model is observed to offer improved recovery compared to current motion-resolved and motion-compensated algorithms for free-breathing 3D cine MRI.

Figures

Figures reproduced from arXiv: 2505.03149 by the authors.

Figure 1
Figure 1. DMoCo architecture: (a) The static template volume [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Impact of low-rank approximation of velocity tensors, deformation fields, and image series: We show the estimated velocity and deformation maps [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 4
Figure 4. Comparison using XCAT phantom: We used the XCAT framework to [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figures from the paper (4 more)
Figure 3
Figure 3. Figure 3: We set the rank of the velocity model to ten. Here, [PITH_FULL_IMAGE:figures/full_fig_p008_3.png]
Figure 5
Figure 5. Figure 5: In-vivo experiments: We show the reconstructions obtained using DMoCO, MoCo-SToRM, and a motion resolved (MoR) approach top. We show the [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]
Figure 6
Figure 6. Figure 6: Reconstructed volumes and segmentations from patient three: We exported the short-axis projections of the 3D data from DMoCo reconstructions. The [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]
Figure 7
Figure 7. Figure 7: Comparison of functional measures with 2D BH cine acquisitions: We compared end-systolic, end-diastolic, and ejection fractions from DMoCo to [PITH_FULL_IMAGE:figures/full_fig_p010_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

48 extracted references · 29 canonical work pages

  1. [23]

    Torres, Sean B

    Qing Zou, Luis A. Torres, Sean B. Fain, Nara S. Higano, 13 Alister J. Bates, and Mathews Jacob. Dynamic imaging using motion-compensated smoothness regularization on manifolds (MoCo-SToRM).Physics in Medicine & Biology, 67(14):144001, 7 2022. ISSN 0031-

    work page 2022

  2. [1]

    Jing Liu, Pascal Spincemaille, Noel C. F. Codella, Thanh D. Nguyen, Martin R. Prince, and Yi Wang. Respiratory and cardiac self-gated free-breathing car- diac cine imaging with multiecho 3d hybrid radial ssfp acquisition.Magnetic Resonance in Medicine, 63(5): 1230–1237, April 2010. ISSN 1522-2594. doi: 10.1002/ mrm.22306. URL http://dx.doi.org/10.1002/mrm.22306

    work page doi:10.1002/mrm.22306 2010

  3. [2]

    Berman, and Debiao Li

    Jianing Pang, Behzad Sharif, Zhaoyang Fan, Xiaoming Bi, Reza Arsanjani, Daniel S. Berman, and Debiao Li. Ecg and navigator-free four-dimensional whole-heart coronary mra for simultaneous visualization of cardiac anatomy and function.Magnetic Resonance in Medicine, 72(5):1208–1217, September 2014. ISSN 1522-2594. doi: 10.1002/mrm.25450. URL http://dx.doi.o...

    work page doi:10.1002/mrm.25450 2014

  4. [3]

    van Heeswijk, Juerg Schwitter, and Matthias Stuber

    Simone Coppo, Davide Piccini, Gabriele Bonanno, J´erˆome Chaptinel, Gabriella Vincenti, H ´el`ene Feliciano, Ruud B. van Heeswijk, Juerg Schwitter, and Matthias Stuber. Free-running 4d whole-heart self-navigated golden angle mri: Initial results.Magnetic Resonance in Medicine, 74(5):1306–1316, November 2014. ISSN 1522-2594. doi: 10.1002/mrm.25523. URL htt...

    work page doi:10.1002/mrm.25523 2014

  5. [4]

    Usman, B

    M. Usman, B. Ruijsink, M.S. Nazir, G. Cruz, and C. Prieto. Free breathing whole-heart 3d cine mri with self-gated cartesian trajectory.Magnetic Resonance Imaging, 38:129–137, May 2017. ISSN 0730-725X. doi: 10.1016/j.mri.2016.12.021. URL http://dx.doi.org/ 10.1016/j.mri.2016.12.021

    doi:10.1016/j.mri.2016.12.021 2017

  6. [5]

    Lim, Pier Giorgio Masci, Matthias Stuber, Daniel K

    Li Feng, Simone Coppo, Davide Piccini, Jerome Yerly, Ruth P. Lim, Pier Giorgio Masci, Matthias Stuber, Daniel K. Sodickson, and Ricardo Otazo. 5d whole- heart sparse mri.Magnetic Resonance in Medicine, 79 (2):826–838, May 2017. ISSN 1522-2594. doi: 10.1002/ mrm.26745. URL http://dx.doi.org/10.1002/mrm.26745

    doi:10.1002/mrm.26745 2017

  7. [6]

    Moghari, Ashita Barthur, Maria E

    Mehdi H. Moghari, Ashita Barthur, Maria E. Amaral, Tal Geva, and Andrew J. Powell. Free-breathing whole-heart 3d cine magnetic resonance imaging with prospective respiratory motion compensation.Magnetic Resonance in Medicine, 80(1):181–189, December 2017. ISSN 1522-2594. doi: 10.1002/mrm.27021. URL http://dx. doi.org/10.1002/mrm.27021

    work page doi:10.1002/mrm.27021 2017

  8. [7]

    An automated approach to fully self-gated free-running cardiac and respiratory motion-resolved 5d whole-heart mri.Mag- netic Resonance in Medicine, 82(6):2118–2132, July 12

    Lorenzo Di Sopra, Davide Piccini, Simone Coppo, Matthias Stuber, and J ´erˆome Yerly. An automated approach to fully self-gated free-running cardiac and respiratory motion-resolved 5d whole-heart mri.Mag- netic Resonance in Medicine, 82(6):2118–2132, July 12

    work page

Show all 48 references
  1. [8]

    Christian M

    Sebastian Rosenzweig, Nick Scholand, H. Christian M. Holme, and Martin Uecker. Cardiac and respiratory self- gating in radial mri using an adapted singular spectrum analysis (ssa-fary).IEEE Transactions on Medical Imag- ing, 39(10):3029–3041, October 2020. ISSN 1558-254X. doi:...

    2020

  2. [9]

    Fully self-gated free-running 3d cartesian cardiac cine with isotropic whole-heart cov- erage in less than 2 min.NMR in Biomedicine, 34(1), September 2020

    Thomas K ¨ustner, Aurelien Bustin, Olivier Jaubert, Reza Hajhosseiny, Pier Giorgio Masci, Radhouene Neji, Ren ´e Botnar, and Claudia Prieto. Fully self-gated free-running 3d cartesian cardiac cine with isotropic whole-heart cov- erage in less than 2 min.NMR in Biomedicine, 34(...

    2020 doi

  3. [10]

    Roy, Bastien Milani, J ´erˆome Yerly, Salim Si-Mohamed, Ludovica Romanin, Aur ´elien Bustin, Es- telle Tenisch, Tobias Rutz, Milan Prsa, and Matthias Stuber

    Christopher W. Roy, Bastien Milani, J ´erˆome Yerly, Salim Si-Mohamed, Ludovica Romanin, Aur ´elien Bustin, Es- telle Tenisch, Tobias Rutz, Milan Prsa, and Matthias Stuber. Intra-bin correction and inter-bin compensation of respiratory motion in free-running five-dimensional w...

  4. [11]

    Arshad, Lee C

    Syed M. Arshad, Lee C. Potter, Chong Chen, Ying- min Liu, Preethi Chandrasekaran, Christopher Crabtree, Matthew S. Tong, Orlando P. Simonetti, Yuchi Han, and Rizwan Ahmad. Motion-robust free-running volumetric cardiovascular mri.Magnetic Resonance in Medicine, May 2024. ISSN 1...

    2024 doi

  5. [12]

    Time- Dependent Deep Image Prior for Dynamic MRI.IEEE Transactions on Medical Imaging, 40(12):3337–3348, 12 2021

    Jaejun Yoo, Kyong Hwan Jin, Harshit Gupta, Jerome Yerly, Matthias Stuber, and Michael Unser. Time- Dependent Deep Image Prior for Dynamic MRI.IEEE Transactions on Medical Imaging, 40(12):3337–3348, 12 2021. ISSN 1558254X. doi: 10.1109/TMI.2021. 3084288

    2021 doi

  6. [13]

    Dynamic imaging using a deep generative SToRM (Gen-SToRM) model.IEEE transactions on medical imaging, 40(11):3102, 11 2021

    Qing Zou, Abdul Haseeb Ahmed, Prashant Nagpal, Stanley Kruger, and Mathews Jacob. Dynamic imaging using a deep generative SToRM (Gen-SToRM) model.IEEE transactions on medical imaging, 40(11):3102, 11 2021. ISSN 1558254X. doi: 10.1109/TMI.2021.3065948. URL /pmc/articles/PMC8590...

    2021

  7. [14]

    Zimmer, Rickmer F

    Veronika Spieker, Wenqi Huang, Hannah Eichhorn, Jonathan Stelter, Kilian Weiss, Veronika A. Zimmer, Rickmer F. Braren, Dimitrios C. Karampinos, Kerstin Hammernik, and Julia A. Schnabel. ICoNIK: Generating Respiratory-Resolved Abdominal MR Reconstructions Using Neural Implicit ...

  8. [15]

    Wenqi Huang, Hongwei Bran Li, Jiazhen Pan, Gastao Cruz, Daniel Rueckert, and Kerstin Hammernik. Neural Implicit k-Space for Binning-free Non-Cartesian Cardiac MR Imaging.Lecture Notes in Computer Science (in- cluding subseries Lecture Notes in Artificial Intelligence and Lectu...

    2022 arXiv

  9. [16]

    Zengyi Li, Yubei Chen, and Friedrich T. Sommer. Learning Energy-Based Models in High-Dimensional Spaces with Multiscale Denoising-Score Matching. Entropy 2023, Vol. 25, Page 1367, 25(10):1367, 9

    2023

  10. [17]

    Schulte, and Mathews Jacob

    Qing Zou, Abdul Haseeb Ahmed, Prashant Nagpal, Sarv Priya, Rolf F. Schulte, and Mathews Jacob. Variational Manifold Learning From Incomplete Data: Application to Multislice Dynamic MRI.IEEE transactions on medical imaging, 41(12):3552–3561, 12 2022. ISSN 1558-254X. doi: 10.110...

    2022

  11. [18]

    Joint reconstruction of multiple images and motion in MRI: Application to free- breathing myocardial quantification.IEEE Transactions on Medical Imaging, 35(1):197–207, 1 2016

    Freddy Odille, Anne Menini, Jean Marie Escany ´e, Pierre Andr ´e Vuissoz, Pierre Yves Marie, Marine Beau- mont, and Jacques Felblinger. Joint reconstruction of multiple images and motion in MRI: Application to free- breathing myocardial quantification.IEEE Transactions on Medi...

    2016

  12. [19]

    Botnar, and Claudia Prieto

    Camila Munoz, Haikun Qi, Gastao Cruz, Thomas K¨ustner, Ren ´e M. Botnar, and Claudia Prieto. Self- supervised learning-based diffeomorphic non-rigid mo- tion estimation for fast motion-compensated coronary MR angiography.Magnetic resonance imaging, 85: 10–18, 1 2022. ISSN 1873...

    2022

  13. [20]

    Fletcher, Haikun Qi, Ren ´e M

    Andrew Phair, Anastasia Fotaki, Lina Felsner, Thomas J. Fletcher, Haikun Qi, Ren ´e M. Botnar, and Claudia Pri- eto. A motion-corrected deep-learning reconstruction framework for accelerating whole-heart magnetic res- onance imaging in patients with congenital heart dis- ease....

    2024

  14. [21]

    doi: 10.3390/E25101367

    ISSN 1099-4300. doi: 10.3390/E25101367. URL https://www.mdpi.com/1099-4300/25/10/1367/htmhttps: //www.mdpi.com/1099-4300/25/10/1367

    doi

  15. [22]

    Un- rolled and rapid motion-compensated reconstruction for cardiac CINE MRI.Medical Image Analysis, 91:103017, 1 2024

    Jiazhen Pan, Manal Hamdi, Wenqi Huang, Kerstin Ham- mernik, Thomas Kuestner, and Daniel Rueckert. Un- rolled and rapid motion-compensated reconstruction for cardiac CINE MRI.Medical Image Analysis, 91:103017, 1 2024. ISSN 1361-8415. doi: 10.1016/J.MEDIA.2023. 103017

    2024 doi

  16. [24]

    Accelerated real-time cine and flow under in-magnet staged exer- cise.Journal of Cardiovascular Magnetic Resonance, 27(1):101894, 2025

    Preethi S Chandrasekaran, Chong Chen, Yingmin Liu, Syed Murtaza Arshad, Christopher Crabtree, Matthew Tong, Yuchi Han, and Rizwan Ahmad. Accelerated real-time cine and flow under in-magnet staged exer- cise.Journal of Cardiovascular Magnetic Resonance, 27(1):101894, 2025. ISSN...

    2025

  17. [25]

    Motion-guided deep image prior for cardiac mri.arxiv, 2025

    Marc V ornehm, Chong Chen, Muhammad Ahmad Sultan, Syed Murtaza Arshad, Yuchi Han, Florian Knoll, and Rizwan Ahmad. Motion-guided deep image prior for cardiac mri.arxiv, 2025. doi: https://doi.org/10.48550/ arXiv.2412.04639

    2025 doi arXiv

  18. [26]

    Johnson, and Peder E

    Xucheng Zhu, Marilynn Chan, Michael Lustig, Kevin M. Johnson, and Peder E. Z. Larson. Iterative motion- compensation reconstruction ultra-short te (imoco ute) for high-resolution free-breathing pulmonary mri.Mag- netic Resonance in Medicine, 83(4):1208–1221, Septem- ber 2019. ...

    2019 doi

  19. [27]

    Sabuncu, John Guttag, and Adrian V

    Guha Balakrishnan, Amy Zhao, Mert R. Sabuncu, John Guttag, and Adrian V . Dalca. V oxelMorph: A Learning Framework for Deformable Medical Image Registration. IEEE Transactions on Medical Imaging, 38(8), 2019. ISSN 1558254X. doi: 10.1109/TMI.2019.2897538

    2019

  20. [28]

    Christensen, Richard D

    Gary E. Christensen, Richard D. Rabbitt, and Michael I. Miller. Deformable templates using large deformation kinematics.IEEE Transactions on Image Processing, 5 (10), 1996. ISSN 10577149. doi: 10.1109/83.536892

    1996 doi

  21. [29]

    Faisal Beg, Michael I

    M. Faisal Beg, Michael I. Miller, Alain Trouv ´e, and Laurent Younes. Computing large deformation metric mappings via geodesic flows of diffeomorphisms.Inter- national Journal of Computer Vision, 61(2), 2005. ISSN 09205691. doi: 10.1023/B:VISI.0000043755.93987.aa

    2005

  22. [30]

    Sodickson, and Ricardo Otazo

    Li Feng, Leon Axel, Hersh Chandarana, Kai Tobias Block, Daniel K. Sodickson, and Ricardo Otazo. Xd- grasp: Golden-angle radial mri with reconstruction of extra motion-state dimensions using compressed sensing. Magnetic Resonance in Medicine, 75(2):775–788, March

  23. [31]

    Davide Piccini, Arne Littmann, Sonia Nielles-Vallespin, and Michael O. Zenge. Spiral phyllotaxis: The natural way to construct a 3d radial trajectory in mri.Magnetic Resonance in Medicine, 66(4):1049–1056, April 2011. ISSN 1522-2594. doi: 10.1002/mrm.22898. URL http: //dx.doi....

    2011 doi

  24. [32]

    Springer Nature Switzerland, 2023

    Joseph Kettelkamp, Ludovica Romanin, Davide Piccini, Sarv Priya, and Mathews Jacob.Motion Compen- sated Unsupervised Deep Learning for 5D MRI, page 419–427. Springer Nature Switzerland, 2023. ISBN 9783031439995. doi: 10.1007/978-3-031-43999-5 40. URL http://dx.doi.org/10.1007/...

    2023 doi

  25. [33]

    Peters, Kay Nehrke, and Warren J

    Reza Nezafat, Daniel Herzka, Christian Stehning, Dana C. Peters, Kay Nehrke, and Warren J. Manning. Inflow quantification in three-dimensional cardiovascular mr imaging.Journal of Magnetic Resonance Imaging, 28(5):1273–1279, October 2008. ISSN 1522-2586. doi: 10.1002/jmri.2149...

    2008 doi

  26. [34]

    Treffer, Michael Weyand, and Oliver Rompel

    Muhnnad Alkassar, Sophia Engelhardt, Tariq Abu-Tair, Efren Ojeda, Philipp C. Treffer, Michael Weyand, and Oliver Rompel. Comparative study of 2d-cine and 3d- wh volumetry: Revealing systemic error of 2d-cine vol- umetry.Diagnostics, 13(20):3162, October 2023. ISSN 2075-4418. d...

    2023 doi

  27. [35]

    Comparison between three-dimensional navigator-gated whole-heart mri and two-dimensional cine mri in quantifying ventricular volumes.Korean Journal of Radiology, 19(4):704, 2018

    Hyun Woo Goo. Comparison between three-dimensional navigator-gated whole-heart mri and two-dimensional cine mri in quantifying ventricular volumes.Korean Journal of Radiology, 19(4):704, 2018. ISSN 2005-8330. doi: 10.3348/kjr.2018.19.4.704. URL http://dx.doi.org/ 10.3348/kjr.2...

    2018 doi

  28. [36]

    Design and validation of segment - freely available software for cardiovascular image analysis.BMC Medical Imag- ing, 10(1), January 2010

    Einar Heiberg, Jane Sj ¨ogren, Martin Ugander, Marcus Carlsson, Henrik Engblom, and H ˚akan Arheden. Design and validation of segment - freely available software for cardiovascular image analysis.BMC Medical Imag- ing, 10(1), January 2010. ISSN 1471-2342. doi: 10. 1186/1471-23...

    2010

  29. [37]

    Bastiaansen, Davide Piccini, Lorenzo Di Sopra, Christopher W

    Jessica A.M. Bastiaansen, Davide Piccini, Lorenzo Di Sopra, Christopher W. Roy, John Heerfordt, Robert R. Edelman, Ioannis Koktzoglou, J ´erˆome Yerly, and Matthias Stuber. Natively fat-suppressed 5d whole-heart mri with a radial free-running fast-interrupted steady- state (fi...

    2019

  30. [38]

    Bashir, Lubna Bhatti, Daniele Marin, and Rendon C

    Mustafa R. Bashir, Lubna Bhatti, Daniele Marin, and Rendon C. Nelson. Emerging applications for ferumoxy- tol as a contrast agent in mri.Journal of Magnetic Resonance Imaging, 41(4):884–898, June 2014. ISSN 1522-2586. doi: 10.1002/jmri.24691. URL http://dx.doi. org/10.1002/jmri.24691

    2014 doi

  31. [39]

    https://cvit.duke

    XCAT Phantom Program - CVIT - Center for Vir- tual Imaging Trials — cvit.duke.edu. https://cvit.duke. edu/resource/xcat-phantom-program/. [Accessed 11-03- 2025]

    2025

  32. [40]

    Multi-scale energy (muse) framework for inverse problems in imag- ing.IEEE Transactions on Computational Imaging, 10: 1250–1265, 2024

    Jyothi Rikhab Chand and Mathews Jacob. Multi-scale energy (muse) framework for inverse problems in imag- ing.IEEE Transactions on Computational Imaging, 10: 1250–1265, 2024. ISSN 2573-0436. doi: 10.1109/tci. 2024.3449101. URL http://dx.doi.org/10.1109/TCI.2024. 3449101

    2024

  33. [47]

    Roy, Lorenzo Di Sopra, Kevin K

    Christopher W. Roy, Lorenzo Di Sopra, Kevin K. Whitehead, Davide Piccini, Me Yerly, John Heerfordt, Reena M. Ghosh, Mark A. Fogel, and Matthias Stuber. Free-running cardiac and respiratory motion-resolved 5d whole-heart coronary cardiovascular magnetic resonance angiography in...

    2022 doi

  34. [2015]

    doi: 10.1002/mrm.25665

    ISSN 1522-2594. doi: 10.1002/mrm.25665. URL http://dx.doi.org/10.1002/mrm.25665

    doi

  35. [2019]

    doi: 10.1002/mrm.27898

    ISSN 1522-2594. doi: 10.1002/mrm.27898. URL http://dx.doi.org/10.1002/mrm.27898

    doi

  36. [2023]

    URL https://arxiv.org/abs/2308.08830v1

    arXiv

  37. [2024]

    doi: 10.1016/j.jocmr.2024

    ISSN 1097-6647. doi: 10.1016/j.jocmr.2024. 101037. URL http://dx.doi.org/10.1016/j.jocmr.2024. 101037

    2024 doi

  38. [2594]

    URL http://dx.doi.org/ 10.1002/mrm.27942

    doi: 10.1002/mrm.27942. URL http://dx.doi.org/ 10.1002/mrm.27942

    doi

  39. [4409]

    URL http://dx.doi.org/10.1002/nbm.4409

    doi

  40. [9155]

    URL https://iopscience.iop.org/article/10.1088/1361-6560/ ac79fchttps://iopscience.iop.org/article/10.1088/ 1361-6560/ac79fc/meta

    doi: 10.1088/1361-6560/AC79FC. URL https://iopscience.iop.org/article/10.1088/1361-6560/ ac79fchttps://iopscience.iop.org/article/10.1088/ 1361-6560/ac79fc/meta

    doi

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.