Pith. sign in

REVIEW 4 major objections 9 minor 28 references

A Novel Coronary Artery Registration Method Based on Super-pixel Particle Swarm Optimization

T0 review · 4 major / 9 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read This paper claims a super-pixel particle swarm optimization pipeline registers preoperative 3D coronary CTA to intraoperative 2D X-ray angiography with an average error of 2.67 mm, outperforming four state-of-the-art baselines.

desk verdict A clinically motivated but methodologically broken registration paper: the claimed SOTA comparison never actually appears in the manuscript. read the letter →

arxiv 2505.24351 v1 pith:SLNSQE5F submitted 2025-05-30 eess.IV cs.CV

Peng Qi , Wenxi Qu , Tianliang Yao , Haonan Ma , Dylan Wintle , Yinyi Lai , Giorgos Papanastasiou , Chengjia Wang This is my paper
classification eess.IVcs.CV
keywords multimodalimageregistrationcoronaryarteryCTangiographyX-rayparticleswarmoptimizationsuper-pixelvesselsegmentationpercutaneousintervention
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

Percutaneous coronary intervention (PCI) is guided in real time by 2D X-ray angiography, which lacks depth, while preoperative CT angiography provides 3D coronary anatomy. This paper proposes a registration method that aligns the two by turning both into weighted discrete point sets and searching for the 3D-to-2D pose with particle swarm optimization. On a pilot dataset of 28 CTA/XRA image pairs from 10 PCI patients, the method reports an average registration error of 2.67 mm measured as minimum projection distance, and the authors claim it beats four competing baseline families in accuracy, robustness, and efficiency. The value, if true, is a practical way to overlay 3D CTA context onto live X-ray images during PCI, potentially improving navigation through occluded or twisted vessels.

What carries the argument

The load-bearing object is the super-pixel (called super-voxel when projected) representation: the vascular tree is reduced to weighted discrete points at edges, centerline locations, and bifurcations, so that registration becomes a particle-swarm search over a six-DOF virtual camera pose. DSA features come from UNet segmentation, K3M skeletonization, Shi-Tomasi corner detection, and moment-based centroid extraction; CTA features come from dynamic threshold segmentation of the 3D vessel tree. The particle swarm updates particle velocities and positions, projects the 3D points to 2D, and uses normalized cross-correlation as the stopping criterion, with weights emphasizing bifurcation points.

What would settle it

Run the SPSO pipeline on a held-out set of clinical CTA/XRA pairs without manually removing CTA segments that have no matching XRA vessel, and measure mPD against expert-annotated ground truth. If the average error rises to the level of the baselines or above, the reported superiority depends on the manual preprocessing rather than on the registration method itself.

Watch

Extended reading notes

Core claim

The paper's central claim is that coronary CTA-to-XRA registration is best handled as a weighted discrete-point matching problem: vessel centerlines, edges, and bifurcations are extracted from both modalities, discretized into 'super-voxels,' and aligned by a particle swarm optimizer that searches the six-degree-of-freedom camera pose and scores each pose by normalized cross-correlation. The authors report that this SPSO pipeline achieves an average registration error of 2.67 mm on the pilot dataset, outperforming the ICP-based, topology-based, Gaussian mixture model-based, and feature point-based baselines compared in Table 1. They also claim that UNet is the best of the six deep segmentation models they tested for the DSA preprocessing module, with a Dice coefficient of 0.7166 and an overall matching accuracy of 0.9779, and that using bifurcation features in addition to centerlines is necessary to avoid matching errors at branch starting points.

Load-bearing premise

The central claim rests on the assumption that an error metric computed on manually trimmed data with known point correspondences predicts how well the method will register real clinical images, where correspondences are unknown and some vessel segments may be occluded or missing.

Editorial extensions

If this is right

  • If the 2.67 mm average mPD transfers to clinical use, the pipeline could furnish real-time 3D CTA overlay during PCI with error below typical coronary vessel diameters.
  • The weighted discrete-point formulation lets the optimizer prioritize bifurcations, so registration accuracy concentrates on the landmarks surgeons use for navigation.
  • Because the method uses simple, fast PSO rather than a learned deformation model, it may be easier to deploy or adapt to new imaging setups without retraining.
  • The UNet segmentation experiment implies that better DSA vessel segmentation directly translates into better registration, making segmentation quality a primary lever for further gains.
  • The reported consistency across left and right coronary arteries suggests the method is not tied to one vascular geometry.

Reading between the lines

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

  • The authors do not test registration on untrimmed CTA data; a natural next experiment is to hold out cases with occluded or stenotic vessels and compare mPD with and without manual CTA trimming. If the gap is large, the clinical advantage depends on automatic handling of missing segments.
  • A direct extension is to replace the PSO search with a differentiable pose estimator trained on the same weighted super-voxel representation, which could remove the need for hand-set PSO parameters and speed up convergence.
  • The same discretize-and-swarm strategy could apply to other 3D/2D vessel registration problems, such as cerebral or peripheral angiography, wherever centerlines and bifurcations can be extracted.
  • Reported mPD values near 2.7 mm should be compared against coronary vessel diameter and clinical tolerance; whether 2.67 mm is adequate for guiding PCI is a clinical question the paper does not answer.
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

4 major / 9 minor

Summary. This paper proposes a 3D-2D coronary artery registration method, SPSO, that combines deep-learning-based DSA segmentation with a particle swarm optimization search over virtual camera poses, using discretized 'super-voxel' point sets extracted from DSA and CTA images. The authors report on a pilot dataset of 28 XRA/CTA pairs from 10 patients, claim that the method outperforms nine state-of-the-art baselines in accuracy, robustness, and efficiency, and report an average registration error of 2.67 mm. Section IV-C describes the baseline comparison protocol, Section III-D defines the mPD metric used for evaluation, and Section IV-D compares six segmentation backbones.

Significance. The clinical problem is relevant and a robust CTA-to-XRA registration method would be valuable for PCI guidance. The paper presents a coherent pipeline: segmentation, centerline extraction, discretization, and PSO-based pose optimization. The inclusion of a deep segmentation network and the use of a swarm optimizer are reasonable engineering choices. However, the central claims are not supported by the evidence presented: no results for the baselines are shown, the evaluation metric is the same objective used for optimization, and the test data are manually pruned to remove mismatches. As submitted, the manuscript does not yet establish that the method performs better than existing approaches, so the significance of the contribution cannot be assessed.

major comments (4)
  1. [Section IV-C, Tables 1-2] The abstract claims that SPSO 'outperformed the selected SOTA baselines in all aspects' of registration accuracy, robustness, and efficiency, and Section IV-C states that the method was compared with nine competing methods (ICP-BP, ICP-PnP, Tree, DT, CS, GMM, OGMM, PSO-EM, Grad). However, no table or figure in the manuscript reports accuracy, robustness, or runtime results for any of these baselines. Table 1 lists only parameter settings, and Table 2 reports mPD values and segmentation metrics for different DSA segmentation backbones (R2Net, ResUNet++, TransResUNet, U2Net, UNet, VGGNet), not for any of the nine registration methods. The only quantitative registration result, the 2.67 mm average in Section V, is the UNet row of Table 2 and has no baseline comparison. The central comparative claim is therefore unsupported and cannot be verified from the submitted manuscript.
  2. [Sections III-D and IV-B] The evaluation protocol is circular: Eq. (13) defines mPD as the average distance from each 3D point to its projection on the 2D plane, and Section IV-B states that 'since the data is generated based on the known point correspondence relationship, mPD is used in the loss function instead of soft dtw when performing experiments on the dataset.' Section III-D then presents mPD as the registration accuracy metric. The reported mPD values, including the headline 2.67 mm average, are therefore values of the very same objective that was minimized during optimization. This is not an independent assessment of registration accuracy, and any conclusion drawn from these numbers about the quality of the registration is invalid.
  3. [Section V] The paper explicitly states that imperfect clinical data are handled by 'artificially removing the corresponding parts in the CTA images to achieve a 1:1 matching effect.' This manual pruning removes the vessel segments that do not have a counterpart in the XRA image, which are precisely the cases where occlusions or mismatches occur in real clinical data. As a result, the reported 'average registration error of the method on the dataset is 2.67mm' is an average over a dataset from which the difficult cases have been eliminated, and it cannot support the claim of robustness on real clinical images with unknown correspondences.
  4. [Section III-D, Eq. (13), and Table 2] The text in Section III-D describes mPD as a 'minimum projected distance': 'for each point in the 3D vascular structure, we project it onto the 2D vascular plane along all possible directions and choose the direction that minimizes the projection distance.' However, Eq. (13) is a simple average of d(a, P_B(a)) over points a in the 3D structure, with no minimization over directions. These two definitions are inconsistent, so the reported numeric values cannot be reproduced. In addition, Table 2 shows that the method's mPD for the right coronary artery (2.66 mm) is lower than the manually annotated ground-truth mPD (2.99 mm), which indicates that mPD as computed is not a true registration error with respect to the manual reference.
minor comments (9)
  1. [Abstract and Section IV-C] The abstract says the method was compared with 'four state-of-the-art (SOTA) methods,' while Section IV-C says it was compared with 'nine competing methods.' Please reconcile the number.
  2. [Section III-B] The section title is 'CTA Image Keypoint Extraction Module,' but the first paragraph describes a 'DSA Image Keypoint Extraction Module.' The heading and the content need to be aligned.
  3. [Sections V and VI] The paper has two consecutive discussion sections, 'Discussions and Conclusions' and 'Discussions and Future Work,' with overlapping content. They should be merged or clearly separated.
  4. [Section IV-C and IV-D] There are placeholders in the text: 'Figure A shows the basic registration results' and 'The registration results of the six competitive methods are shown in Figure XXX.' These need to be replaced with actual figure references.
  5. [Table 2] The table title 'PARAMETERS OF 3D-2D REGISTRATION METHODS' is incorrect; the table reports mPD values and segmentation performance metrics, not method parameters.
  6. [Eq. (13) and Section III-D] The notation in Eq. (13) is underspecified: d is not defined as a distance metric, and P_B(a) is called a projection onto plane B even though B is defined as a set of points in the 2D vascular plane. Please clarify the mathematical objects.
  7. [Section I, Contributions] The list of contributions contains duplicated bullet markers ('• •') on two items.
  8. [Throughout] The terms 'super-pixel,' 'super-voxel,' 'hypergraph,' and 'hyperedge' are used interchangeably. Choose one consistent terminology and define it at first use.
  9. [Section IV-A] The dataset section says 1571 XRA images were annotated and split 8:1:1, while Section IV-D mentions training on '968 pairs of DSA images.' Please clarify the relationship between these numbers.

Circularity Check

1 steps flagged · score 6.0 of 10

Reported registration error is the optimized mPD loss; SOTA comparisons are not reported.

  1. fitted input called prediction [Section IV-B (Implementation Details); Section III-D (Evaluation Metrics); Section V (Discussions and Conclusions)]
    "Since the data is generated based on the known point correspondence relationship, mPD is used in the loss function instead of soft dtw when performing experiments on the dataset. ... To evaluate the performance of the proposed registration method, we use the minimum projected distance (mPD) [19] as a metric. ... The average registration error of the method on the dataset is 2.67mm."

    The pose parameters are optimized by minimizing the mPD objective (Section IV-B), and the same mPD is then defined as the evaluation metric (Section III-D) and reported as the headline accuracy (Section V). The reported 2.67mm is therefore the value of the minimized cost function at the fitted pose, not an independent measure of registration accuracy on unseen data. Minimizing a function and then reporting that function's value as 'error' does not test generalization; it is circular by construction. Table II confirms that the claimed error is simply the UNet row's mPD value, and no baseline mPD values are provided, so the comparative claim cannot be independently verified.

full rationale

The paper's central quantitative result, 'The average registration error of the method on the dataset is 2.67mm', reduces to the value of the mPD objective that the particle swarm optimizer was explicitly run to minimize. Since Section IV-B states that 'mPD is used in the loss function' and Section III-D defines mPD as the evaluation metric, the reported registration error is the optimized loss, not an independent prediction. This is a genuine circular step. Separately, the abstract claims the method 'outperformed the selected SOTA baselines in all aspects', but Section IV-C lists nine competing methods and no table or figure reports their registration accuracy, robustness, or runtime; only segmentation-backbone comparisons appear in Table II. That absence is an evidence gap rather than a circularity, but it further weakens the headline claim. No load-bearing self-citation or imported uniqueness theorem was found. Because the main numerical performance claim is forced by the choice of the same mPD as loss and metric, a moderate-to-substantial circularity score of 6 is appropriate.

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

The central claim rests on a trained segmentation network, user-tuned PSO and threshold parameters, and multiple domain assumptions about data validity. No new physical entities are introduced.

free parameters (4)
  • NCC threshold e = not reported
    User-defined stopping threshold 'e' in Section III-C, 'determined through experimental exploration and practical requirements'.
  • PSO hyperparameters (omega, c1, c2, population size) = not reported
    Velocity update in Eq. 10 requires inertia weight omega and learning factors c1, c2; values are not specified, only max iterations 400 in Table 1.
  • Point weights for weighted similarity = not reported
    Section V assigns weights to discrete points with focus on bifurcations; the weighting scheme is described qualitatively and chosen by hand.
  • UNet segmentation weights = trained model
    The DSA segmentation network is trained on annotated images (Section IV-A); the trained model is a fitted function that the registration depends on.
assumptions (4)
  • domain assumption A six-degree-of-freedom rigid camera pose can represent the CTA-to-XRA transformation despite cardiac and respiratory deformation.
    Section III-C adjusts only three rotations and one translation vector (6 DoF), ignoring nonrigid deformation.
  • ad hoc to paper Manually removing mismatching CTA vessel segments to achieve 1:1 matching is a valid preprocessing step.
    Section V: 'we handle these imperfect data by artificially removing the corresponding parts in the CTA images to achieve a 1:1 matching effect.'
  • domain assumption mPD computed with known point correspondences is a valid measure of registration accuracy on clinical data.
    Section IV-B uses mPD as loss because 'the data is generated based on the known point correspondence relationship'; this assumes such a metric transfers to unseen clinical images without correspondences.
  • standard math Standard PSO update equations and K3M skeletonization from prior literature are correct.
    Eqs. 10-11 and Eq. 1-2 are taken from refs [18] and [15] without proof.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Novel Coronary Artery Registration Method Based on Super-pixel Particle Swarm Optimization." pith.science (2026). https://pith.science/paper/SLNSQE5F

@misc{pith2026250524351,
  author       = {Pith},
  title        = {Pith review of: A Novel Coronary Artery Registration Method Based on Super-pixel Particle Swarm Optimization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SLNSQE5F}},
  note         = {Machine review of arXiv:2505.24351}
}
read the original abstract

Percutaneous Coronary Intervention (PCI) is a minimally invasive procedure that improves coronary blood flow and treats coronary artery disease. Although PCI typically requires 2D X-ray angiography (XRA) to guide catheter placement at real-time, computed tomography angiography (CTA) may substantially improve PCI by providing precise information of 3D vascular anatomy and status. To leverage real-time XRA and detailed 3D CTA anatomy for PCI, accurate multimodal image registration of XRA and CTA is required, to guide the procedure and avoid complications. This is a challenging process as it requires registration of images from different geometrical modalities (2D -> 3D and vice versa), with variations in contrast and noise levels. In this paper, we propose a novel multimodal coronary artery image registration method based on a swarm optimization algorithm, which effectively addresses challenges such as large deformations, low contrast, and noise across these imaging modalities. Our algorithm consists of two main modules: 1) preprocessing of XRA and CTA images separately, and 2) a registration module based on feature extraction using the Steger and Superpixel Particle Swarm Optimization algorithms. Our technique was evaluated on a pilot dataset of 28 pairs of XRA and CTA images from 10 patients who underwent PCI. The algorithm was compared with four state-of-the-art (SOTA) methods in terms of registration accuracy, robustness, and efficiency. Our method outperformed the selected SOTA baselines in all aspects. Experimental results demonstrate the significant effectiveness of our algorithm, surpassing the previous benchmarks and proposes a novel clinical approach that can potentially have merit for improving patient outcomes in coronary artery disease.

Figures

Figures reproduced from arXiv: 2505.24351 by the authors.

Figure 1
Figure 1. Workflow of the Coronary Artery Registration [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Vessel segmentation of DSA images is performed using the UNet deep learning model [14]. UNet is a commonly used semantic segmentation model that is well-suited for medical image segmentation tasks. The UNet model consists of an encoder and a decoder, enabling effective extraction of semantic features from the image and restoration of spatial resolution. The encoder utilizes the Conv class as the basic convolutional … view at source ↗
Figure 3
Figure 3. K3M Skeleton Extraction(schematic diagram) [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (4 more)
Figure 2
Figure 2. Figure 2: Illustration of the Segmentation Network Structure [PITH_FULL_IMAGE:figures/full_fig_p004_2.png]
Figure 5
Figure 5. Figure 5: Particle Swarm Optimization(PSO) For this purpose, we adjust the six degrees of freedom of the virtual camera (consisting of three rotation angles and one translation vector). After each adjustment of the virtual camera’s pose, we project the 3D super-voxels onto the 2…
Figure 6
Figure 6. Figure 6: Image registration result flow [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: registration results of L and R coronary arteries [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

28 extracted references · 28 canonical work pages

  1. [1]

    Percutaneous coronary intervention in 2018,

    D. L. Bhatt, “Percutaneous coronary intervention in 2018,” JAMA, vol. 319, no. 20, p. 2127–2128, 2018

    work page 2018

  2. [2]

    T. M. Peters and K. Cleary, eds., Image-guided Interventions: Technol- ogy and Applications . Springer, 2008

    work page 2008

  3. [3]

    A novel 2d-3d registration algorithm for aligning fluoro images with 3d pre-op ct/mr images,

    H. Sundar, A. Khamene, C. Xu, F. Sauer, and C. Davatzikos, “A novel 2d-3d registration algorithm for aligning fluoro images with 3d pre-op ct/mr images,” in Proceedings of SPIE - The International Society for Optical Engineering , vol. 6141, pp. 760–766, 2006

    work page 2006

  4. [4]

    A statistical model of catheter motion from interventional x-ray images: Application to image-based gating,

    M. Panayiotou, A. King, Y . Ma, R. Housden, C. Rinaldi, J. Gill, M. Cooklin, M. O’Neill, and K. Rhode, “A statistical model of catheter motion from interventional x-ray images: Application to image-based gating,” Physics in Medicine and Biology , vol. 58, no. 21, p. 7543, 2013

    work page 2013

  5. [5]

    Real-time respiratory signal extraction from x-ray sequences using incremental manifold learning,

    P. Fischer, T. Pohl, A. Faranesh, A. Maier, and J. Hornegger, “Real-time respiratory signal extraction from x-ray sequences using incremental manifold learning,” in IEEE International Symposium on Biomedical Imaging, pp. 915–918, 2014

    work page 2014

  6. [6]

    Pca- derived respiratory motion surrogates from x-ray angiograms for per- cutaneous coronary interventions,

    H. Ma, G. Dibildox, C. Schultz, E. Regar, and T. van Walsum, “Pca- derived respiratory motion surrogates from x-ray angiograms for per- cutaneous coronary interventions,” International Journal of Computer Assisted Radiology and Surgery , vol. 10, no. 5, pp. 695–705, 2015

    work page 2015

  7. [7]

    Tracking vessels in x-ray angiogram sequences based on deformable model,

    S. Zheng and Y . Daoyin, “Tracking vessels in x-ray angiogram sequences based on deformable model,” Iranian Journal of Electrical and Com- puter Engineering , vol. 4, no. 2, pp. 137–143, 2005

    work page 2005

  8. [8]

    Vessel tree tracking in angiographic sequences,

    D. Zhang, S. Sun, Z. Wu, B.-J. Chen, and T. Chen, “Vessel tree tracking in angiographic sequences,” Journal of Medical Imaging , vol. 4, no. 2, p. 025001, 2017

    work page 2017

Show all 28 references
  1. [9]

    Extraction of coronary vessels in fluoroscopic x-ray sequences using vessel corre- spondence optimization,

    S. Y . Shin, S. Lee, K. J. Noh, I. D. Yun, and K. M. Lee, “Extraction of coronary vessels in fluoroscopic x-ray sequences using vessel corre- spondence optimization,” in Medical Image Computing and Computer- Assisted Intervention - MICCAI 2016 , p. 308–316, Springer, 2016

    2016

  2. [10]

    Curvilinear structure tracking by low rank tensor approximation with model propagation,

    E. Cheng, Y . Pang, Y . Zhu, J. Yu, and H. Ling, “Curvilinear structure tracking by low rank tensor approximation with model propagation,” in IEEE Conference on Computer Vision and Pattern Recognition , p. 3057–3064, 2014

    2014

  3. [11]

    Structure- aware rank-1 tensor approximation for curvilinear structure tracking using learned hierarchical features,

    P. Chu, Y . Pang, E. Cheng, Y . Zhu, Y . Zheng, and H. Ling, “Structure- aware rank-1 tensor approximation for curvilinear structure tracking using learned hierarchical features,” in Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 , p. 413–421, 2016

    2016

  4. [12]

    Intrinsic 2d/3d registration based on a hybrid approach: use in the radiosurgical imaging process,

    M. Vermandel, N. Betrouni, J. Y . Gauvrit, D. Pasquier, C. Vasseur, and J. Rousseau, “Intrinsic 2d/3d registration based on a hybrid approach: use in the radiosurgical imaging process,” Cellular and Molecular Biology, vol. 52, no. 6, p. 44–53, 2006

    2006

  5. [13]

    3d-2d registration of cerebral angiograms: a method and evaluation on clinical images,

    U. Mitrovic, Z. Spiclin, B. Likar, and F. Pernus, “3d-2d registration of cerebral angiograms: a method and evaluation on clinical images,” IEEE Transactions on Medical Imaging , vol. 32, no. 8, p. 1550–1563, 2013

    2013

  6. [14]

    U-net: Convolutional networks for biomedical image segmentation,

    O. Ronneberger, P. Fischer, and T. Brox, “U-net: Convolutional networks for biomedical image segmentation,” Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015 , pp. 234–241, 2015

    2015

  7. [15]

    K3m: A universal algorithm for image skeletonization and a review of thinning tech- niques,

    K. Saeed, M. Tabedzki, M. Rybnik, and M. Adamski, “K3m: A universal algorithm for image skeletonization and a review of thinning tech- niques,” International Journal of Applied Mathematics and Computer Science, vol. 20, no. 2, pp. 317–335, 2010

    2010

  8. [16]

    Materialise mimics | 3d medical image processing soft- ware: www.materialise.com,

    Materialise, “Materialise mimics | 3d medical image processing soft- ware: www.materialise.com,” in Materialise Mimics | 3D Medical Image Processing Software: www.materialise.com , Year

  9. [17]

    Cpr-gcn: Condi- tional partial-residual graph convolutional network in automated anatom- ical labeling of coronary arteries,

    H. Yang, X. Zhen, Y . Chi, L. Zhang, and X.-S. Hua, “Cpr-gcn: Condi- tional partial-residual graph convolutional network in automated anatom- ical labeling of coronary arteries,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pp. 386–395, 2020

    2020

  10. [18]

    Particle swarm optimization: developments, applications and resources,

    R. Eberhart and Y . Shi, “Particle swarm optimization: developments, applications and resources,” in Proceedings of the 2001 Congress on Evolutionary Computation , vol. 1, IEEE, 2001

    2001

  11. [19]

    A minimum projected-distance test for parametric single-index berkson models,

    C. Xie and L. Zhu, “A minimum projected-distance test for parametric single-index berkson models,” TEST, vol. 27, no. 3, pp. 700–715, 2018

    2018

  12. [20]

    A tree-topology preserv- ing pairing for 3d/2d registration,

    T. Benseghir, G. Malandain, and R. Vaillant, “A tree-topology preserv- ing pairing for 3d/2d registration,” International Journal of Computer Assisted Radiology and Surgery , vol. 10, no. 6, pp. 913–923, 2015

    2015

  13. [21]

    3d-2d registration of cerebral angiograms: A method and evaluation on clinical images,

    U. Mitrovic, Z. Spiclin, B. Likar, and F. Pernus, “3d-2d registration of cerebral angiograms: A method and evaluation on clinical images,” IEEE Transactions on Medical Imaging , vol. 32, no. 8, pp. 1550–1563, 2013

    2013

  14. [22]

    Iterative closest curve: A framework for curvilinear structure registration application to 2d/3d coronary arteries registration,

    T. Benseghir, G. Malandain, and R. Vaillant, “Iterative closest curve: A framework for curvilinear structure registration application to 2d/3d coronary arteries registration,” in Medical Image Computing and Computer-Assisted Intervention – MICCAI 2013 (K. Mori, I. Sakuma, Y . ...

    2013

  15. [23]

    Robust point set registration using gaussian mixture models,

    B. Jian and B. C. Vemuri, “Robust point set registration using gaussian mixture models,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 33, no. 8, pp. 1633–1645, 2011

    2011

  16. [24]

    Oriented gaussian mixture models for nonrigid2d/3d coronary artery registration,

    N. Baka, C. T. Metz, C. J. Schultz, R.-J. M. van Geuns, W. J. Niessen, and T. van Walsum, “Oriented gaussian mixture models for nonrigid2d/3d coronary artery registration,” IEEE Transactions on Med- ical Imaging , vol. 33, no. 5, pp. 1023–1034, 2014

    2014

  17. [25]

    Nonrigid 2d/3d regis- tration of coronary artery models with live fluoroscopy for guidance of cardiac interventions,

    D. Rivest-Henault, H. Sundar, and M. Cheriet, “Nonrigid 2d/3d regis- tration of coronary artery models with live fluoroscopy for guidance of cardiac interventions,” IEEE Transactions on Medical Imaging , vol. 31, no. 8, pp. 1557–1572, 2012

    2012

  18. [26]

    Robustness and accuracy of feature-based single image 2-d–3-d registration without correspondences for image-guided intervention,

    X. Kang, M. Armand, Y . Otake, W.-P. Yau, P. Y . Cheung, Y . Hu, and R. H. Taylor, “Robustness and accuracy of feature-based single image 2-d–3-d registration without correspondences for image-guided intervention,” IEEE Transactions on Biomedical Engineering , vol. 61, no. 1, ...

    2014

  19. [27]

    Car-net: a deep learning-based deformation model for 3d/2d coronary artery registration,

    W. Wu, J. Zhang, W. Peng, H. Xie, S. Zhang, and L. Gu, “Car-net: a deep learning-based deformation model for 3d/2d coronary artery registration,” IEEE Transactions on Medical Imaging , vol. 41, no. 10, p. 2715–2727, 2022

    2022

  20. [28]

    Transfer learning for nonrigid 2d/3d cardiovascular images registration,

    S. Guan, T. Wang, K. Sun, and C. Meng, “Transfer learning for nonrigid 2d/3d cardiovascular images registration,” IEEE Journal of Biomedical and Health Informatics , vol. 25, no. 9, p. 3300–3309, 2021

    2021

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.