Pith. sign in

REVIEW 4 major objections 5 minor 29 references

Real-Time 3D Guidewire Reconstruction from Intraoperative DSA Images for Robot-Assisted Endovascular Interventions

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read By registering a preoperative 3D CT angiography model to live 2D DSA images and lifting guidewire pixels along back-projected rays with a closed-form depth formula, this paper claims real-time 3D guidewire reconstruction at 39.3 frames…

desk verdict The paper builds a real phantom system and reports good segmentation, but its central 3D reconstruction step is geometrically unfounded and the evaluation metrics don't measure what they claim. read the letter →

arxiv 2506.21631 v1 pith:7N4PV35Z submitted 2025-06-25 cs.RO

Tianliang Yao , Bingrui Li , Bo Lu , Zhiqiang Pei , Yixuan Yuan , Peng Qi This is my paper
classification cs.RO
keywords guidewirereconstruction2D-3DregistrationdigitalsubtractionangiographyCTendovascularroboticsinverseprojectionreal-timeshapesensing
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 sets out to recover the full 3D shape of an interventional guidewire during robot-assisted endovascular procedures, where the only live images are single-view 2D X-ray angiograms (DSA) that carry no depth. The proposed framework first segments the guidewire in the 2D image, then registers the pre-scanned 3D vessel model (CTA) to the live 2D vessel projection by optimizing a rigid transform on SE(3), and finally lifts each 2D guidewire pixel to a 3D point using an inverse-projection formula whose per-point depth is given in closed form. The authors report that this runs in real time at 39.3 frames per second, with a projection error of 1.76 pixels and a length deviation of 2.93 percent, outperforming EPnP and B-spline baselines. If correct, the method would give robotic controllers continuous 3D shape feedback from standard imaging hardware, without adding sensors to the thin guidewire.

What carries the argument

The load-bearing object is the inverse-projection step that converts 2D guidewire pixels into 3D coordinates: $G_{3D} = \{[R^{\top} \mid -R^{\top}t]\,[d_m K^{-1}g_m;1]\}_{m=1}^M$, with the depth $d_m$ given by the closed-form linear-triangulation quotient. This is what turns a segmented 2D skeleton into a 3D curve, and it is the mechanism that makes the reconstruction real-time because each point costs $\mathcal{O}(1)$ arithmetic. The other essential component is the registration stage that supplies $T$: a weighted perspective-$n$-point problem solved on the Lie algebra $\mathfrak{se}(3)$ by manifold-aware Levenberg-Marquardt with a pseudo-Huber robust kernel, which aligns the CTA centerline with the DSA centerline and thus provides the spatial reference frame in which the inverse projection operates.

What would settle it

Image a phantom with a guidewire whose true 3D shape is known from a CT scan, using one DSA view, run the pipeline, and compare the reconstructed depth (distance from the X-ray source) of each point with the known value; large depth errors alongside small 2D projection errors would show that the single-view depth formula is not measuring true 3D position.

Watch

Extended reading notes

Core claim

The central claim is that, once the preoperative 3D CTA centerline is registered to the intraoperative 2D DSA centerline, the 3D guidewire is recovered by a direct linear mapping: each 2D guidewire point $g_m$ is back-projected along its camera ray and placed at depth $d_m = \frac{\|t \times R^{\top}K^{-1}[g_m;1]\|}{\|(R^{\top}K^{-1}[g_m;1]) \times (K^{-1}[g_m;1])\|}$, giving $P_m = T^{-1}(d_m K^{-1}[g_m;1])$ with $T=(R,t)$ the registered rigid pose and $K$ the intrinsic matrix. The authors call this a physics-constrained inverse projection and a form of linear triangulation; it assigns each point a depth in $\mathcal{O}(1)$ time using only the pose, the intrinsics, and the pixel coordinate. On their phantom experiments the reconstruction reaches a projection error of $1.76\pm0.08$ pixels, a length deviation of $2.93\pm0.15\%$, and a frame rate of $39.3\pm1.5$ FPS, and the paper interprets these numbers as evidence that the method provides accurate, real-time 3D spatial information for robotic navigation.

Load-bearing premise

The method assumes that a single 2D X-ray picture, together with the known position and angle of the X-ray source and the alignment between the pre-scanned vessel model and the live picture, uniquely fixes how far each guidewire point is from the X-ray source; without a second picture or a constraint from the vessel wall, that distance is free.

Editorial extensions

If this is right

  • A robotic guidewire manipulator would receive continuous 3D shape feedback at about 39 frames per second from the standard single-view DSA images already used in the operating room, with no sensors added to the guidewire itself.
  • At vascular bifurcations the reconstructed 3D configuration would supply the spatial information needed to decide which branch the wire enters, addressing a decision that 2D images leave ambiguous.
  • The reported projection error ($1.76\pm0.08$ px) and length deviation ($2.93\pm0.15\%$) are lower than the compared EPnP and B-spline baselines, so the framework is claimed to be both more accurate and faster than those methods.
  • Because the reconstruction is a closed-form linear mapping from pixels plus a registered pose, it can be embedded into planning, simulation, and reinforcement-learning pipelines, including the digital-twin direction named in the paper's future work.

Reading between the lines

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

  • The paper states in its evaluation section that no 3D ground truth exists for this problem; projection error and length deviation therefore verify 2D consistency and arc-length preservation rather than true depth accuracy, and a phantom with known 3D positions would be the direct check.
  • The closed-form depth uses only the registered pose and the pixel coordinate, so the vessel model enters the reconstruction only through the alignment $T$; a testable consequence is that depth accuracy should degrade when the live vessel shape deviates from the preoperative CTA.
  • Because each reconstructed point costs $\mathcal{O}(1)$ arithmetic, the pipeline could plausibly be extended to multiple instruments or to dense catheter body reconstruction without a redesign, an extension the paper does not make.
  • Adding an occasional second X-ray view would provide a direct calibration of the monoplane depths and would test whether the single-view assumption is sufficient.
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 / 5 minor

Summary. The paper proposes a framework for reconstructing the 3D shape of a guidewire from intraoperative monoplane 2D DSA images, guided by preoperative 3D CTA. The pipeline consists of segmentation of the guidewire in DSA, extraction of vascular centerlines in both modalities, deformable 3D-2D registration on SE(3), and a so-called inverse projection step that assigns a depth to each 2D guidewire point. The authors report a projection error of 1.76 ± 0.08 px, a length deviation of 2.93 ± 0.15%, and a processing speed of 39.3 ± 1.5 FPS, and claim that the method provides accurate real-time 3D spatial information for robot-assisted endovascular interventions.

Significance. If the 3D reconstruction claim were valid, a monoplane DSA + CTA fusion running at 39 FPS would be a useful contribution to robotic endovascular navigation, where depth information is otherwise missing. The paper has some credible components: a clinical DSA training set with expert annotations, a comparative segmentation study, and a realistic vascular phantom setup. However, the central geometric derivation is not sound: the depth formula in Section II-D does not follow from the imaging model, and the reported accuracy metrics do not measure true 3D accuracy. The central claim of accurate 3D spatial information is therefore unsupported by the evidence presented.

major comments (4)
  1. [Section II-D, depth formula] The closed-form depth d_m is not a valid monoplane triangulation. For a calibrated pinhole camera, a pixel g_m defines a ray {λ K^{-1}[g_m;1] : λ > 0} in camera coordinates, and every λ yields a 3D point P = T^{-1}(λ K^{-1}[g_m;1]) that projects exactly to g_m. The proposed d_m selects the λ that makes d_m K^{-1}[g_m;1] lie on the line t + μ R^T K^{-1}[g_m;1], but nothing in the imaging geometry forces the guidewire to lie on that line, and the CTA centerline C_3D does not appear in the formula. Thus the reconstructed coordinates are an arbitrary choice per ray, not a measurement of the actual guidewire depth.
  2. [Section II-D and Section III-A] The depth formula is singular at the reported initialization. In Section III-A the registration is initialized with R = I and t = [0,0,1000]^T in pixel coordinates. For these values, a = R^T K^{-1}[g;1] equals b = K^{-1}[g;1], so the denominator a × b is zero and d_m is undefined for every guidewire pixel. The formula therefore cannot be evaluated at the exact configuration from which the reported experiments start.
  3. [Section III-C, Eq. (5)] The projection consistency error is a self-consistency metric, not a 3D accuracy metric. Since each reconstructed point is defined as P_m = T^{-1}(d_m K^{-1}[g_m;1]) using the same pixel g_m, applying the projection operator necessarily returns g_m (up to numerical precision). The reported 1.76 ± 0.08 px therefore reflects only segmentation and numerical effects, and would be essentially unchanged for arbitrarily wrong depths. The paper itself states in Section III-C that there is no 3D ground truth, so this metric cannot support the claim of accurate 3D reconstruction.
  4. [Table II and Section III-C] The length deviation metric is not defined. It is unclear whether the reconstructed 3D arc length is compared to the 2D segmentation length, to a known physical guidewire length, or to some other reference. If it is compared to the 2D length, the metric is scale-dependent and can be made small by choosing depth values appropriately; if it is compared to a physical length, that ground truth is not described. A precise definition and an independent 3D ground truth are required before any accuracy claim can be made.
minor comments (5)
  1. [Section III-A] The heading contains a typo: "Experimal Setup" should be "Experimental Setup".
  2. [Figure 4 and Section III-D] The text refers to Fig. 4(c) for both segmentation and reconstruction results, but the caption describes panels (a)-(d); the figure references should be corrected and made consistent with the text.
  3. [Section II-C, Eq. (2)] The covariance matrices Σ_j and Λ_i are introduced without definitions, and the dimensions of Λ_i are not specified consistently with the residual term T^{-1}(π^{-1}(q_j)) - p_i, which is a 3D vector.
  4. [Section II-B] The graph Laplacian and spectral decomposition developed in Section II-B are not used anywhere in the registration or reconstruction pipeline; this section is disconnected from the rest of the method.
  5. [Section II-D] The abstract and discussion describe the reconstruction as "physics-constrained," but no physics-based constraint appears in Section II-D; the depth assignment is purely algebraic and unconstrained.

Circularity Check

1 steps flagged · score 6.0 of 10

Projection-consistency metric is self-consistency by construction; 3D accuracy is validated only against the same 2D pixels used to build the reconstruction.

  1. self definitional [Section II-D (Guidewire Spatial Configuration Reconstruction) and Section III-C, Eq. (5)]
    "'the 3D guidewire coordinates G3D = {P m ∈ R 3}M m=1 are recovered via P m = T −1( d m K −1 [g m; 1] )' and 'The projection consistency error serves as a primary measure of alignment between the reconstructed 3D guidewire and its 2D projections. This metric is defined as: Eproj = 1/N Σ N i=1 ∥P(X3D)−Xi2D∥2 (5)'"

    The reconstructed point P_m is defined by inverse projection of the pixel g_m. For any admissible depth d_m, the perspective projection of P_m is g_m exactly. Thus Eq. (5) re-projects points built from the i-th 2D segmentation onto that same segmentation. The near-zero projection error is a definitional identity, not a test of 3D fidelity: an arbitrarily wrong depth on the ray yields the same error. The claimed 1.76±0.08 px therefore reduces by construction to self-consistency and cannot support the paper's conclusion of accurate 3D spatial information.

full rationale

The quantitative support for the central 3D claim is circular. Section III-C states 'Due to the inherent absence of ground truth data for the 3D guidewire reconstruction problem,' and the only accuracy metric used, Eq. (5), compares the re-projection of reconstructed 3D points with the 2D pixels from which those points were constructed in Section II-D. Since every point on a camera ray projects to the same pixel, projection consistency cannot measure depth or 3D shape; it is satisfied by construction. This warrants partial circularity (6), not full circularity, because the segmentation and registration components are evaluated separately and the paper contains no load-bearing self-citation chain. The separate geometric issue—that a single monoplane ray does not determine guidewire depth—is a correctness defect, not a circularity, but it reinforces that the self-consistency metric is the only support for the 3D reconstruction claim.

Assumptions & free parameters 10 free parameters · 5 assumptions · 0 invented entities

The central reconstruction rests on an unstated and false depth prior: that single-view depth can be recovered from a rigid transform alone. The method also relies on several hand-set thresholds, an assumed camera calibration, and a rigid-body assumption on the vasculature. No new physical entities are introduced.

free parameters (10)
  • centerline ridge threshold epsilon = 0.5
    Used in Section II-A to detect ridge points of the distance transform; chosen empirically.
  • Hessian eigenvalue threshold delta = 1.0
    Used in Section II-A to filter ridge points; chosen empirically.
  • registration smoothness weight lambda = 0.1
    Balances bidirectional reprojection terms in Section II-C; hand-set.
  • feature similarity bandwidth sigma_g = 0.85
    Controls gradient similarity in weighted adjacency matrix in Section II-B; hand-set.
  • geometric uncertainty sigma_geo = 2.0 pixels
    Used in uncertainty propagation in the introduction and Experiments; chosen manually.
  • image uncertainty sigma_img = 0.1
    Used in uncertainty propagation; chosen manually.
  • pseudo-Huber cutoff delta = 1.0
    Used in robust kernel in Section II-C; hand-set.
  • depth range = [100, 1000] pixels
    Constraints on depth estimation in experiments; chosen by hand.
  • initial translation t = [0,0,1000] pixels
    Initializes rigid registration and directly affects the scale of reconstructed depths; a hand-set value that can influence length deviation.
  • camera intrinsics fx, fy, cx, cy = 3000, 3000, 256, 256
    Assumed rather than calibrated from the C-arm; if incorrect, all reconstructed depths and lengths scale accordingly.
assumptions (5)
  • ad hoc to paper A single monoplane 2D projection plus a known rigid transform T uniquely determines the depth of each guidewire point.
    Introduced in Section II-D via the closed-form depth formula; this is geometrically false without a second view or a strong vessel geometry prior.
  • domain assumption The guidewire lies on or near the registered CTA vascular centerline, so its depth can be inherited from the CTA.
    This prior is needed to make monoplane reconstruction well-posed, but it is never stated, used in the depth formula, or validated.
  • domain assumption The phantom vasculature is rigidly fixed and the CTA-to-DSA registration remains valid across all frames.
    Section II-C uses a rigid SE(3) registration; cardiac and respiratory motion are deferred to future work in the Discussion.
  • domain assumption Expert pixel-level annotations with inter-annotator 0.94 are a correct guidewire segmentation ground truth.
    Standard but not independently verified against guidewire artifacts in DSA.
  • standard math Levenberg-Marquardt on the SE(3) manifold converges to a useful optimum.
    Standard local optimization; no global convergence guarantee is provided in Section II-C.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Real-Time 3D Guidewire Reconstruction from Intraoperative DSA Images for Robot-Assisted Endovascular Interventions." pith.science (2026). https://pith.science/paper/7N4PV35Z

@misc{pith2026250621631,
  author       = {Pith},
  title        = {Pith review of: Real-Time 3D Guidewire Reconstruction from Intraoperative DSA Images for Robot-Assisted Endovascular Interventions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7N4PV35Z}},
  note         = {Machine review of arXiv:2506.21631}
}
abstract

Accurate three-dimensional (3D) reconstruction of guidewire shapes is crucial for precise navigation in robot-assisted endovascular interventions. Conventional 2D Digital Subtraction Angiography (DSA) is limited by the absence of depth information, leading to spatial ambiguities that hinder reliable guidewire shape sensing. This paper introduces a novel multimodal framework for real-time 3D guidewire reconstruction, combining preoperative 3D Computed Tomography Angiography (CTA) with intraoperative 2D DSA images. The method utilizes robust feature extraction to address noise and distortion in 2D DSA data, followed by deformable image registration to align the 2D projections with the 3D CTA model. Subsequently, the inverse projection algorithm reconstructs the 3D guidewire shape, providing real-time, accurate spatial information. This framework significantly enhances spatial awareness for robotic-assisted endovascular procedures, effectively bridging the gap between preoperative planning and intraoperative execution. The system demonstrates notable improvements in real-time processing speed, reconstruction accuracy, and computational efficiency. The proposed method achieves a projection error of 1.76$\pm$0.08 pixels and a length deviation of 2.93$\pm$0.15\%, with a frame rate of 39.3$\pm$1.5 frames per second (FPS). These advancements have the potential to optimize robotic performance and increase the precision of complex endovascular interventions, ultimately contributing to better clinical outcomes.

Figures

Figures reproduced from arXiv: 2506.21631 by the authors.

Figure 1
Figure 1. The robotic-assisted endovascular procedure is a minimally invasive [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Overview of robotic-assisted endovascular procedures with compu [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Overview of our proposed guidewire 3D reconstruction framework. (a) The pipeline consists of multiple key components: CTA and DSA image [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Overview of the robotic system and experimental results for 3D spatial reconstruction of a guidewire. (a) The hardware setup of the robotic [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

29 extracted references · 24 canonical work pages

  1. [1]

    Autonomous navigation for robot-assisted intraluminal and en- dovascular procedures: A systematic review,

    A. Pore, Z. Li, D. Dall’Alba, A. Hernansanz, E. De Momi, A. Men- ciassi, A. C. Gelpi, J. Dankelman, P. Fiorini, and E. Vander Poorten, “Autonomous navigation for robot-assisted intraluminal and en- dovascular procedures: A systematic review,”IEEE Transactions on Robotics, vol. 39, no. 4, pp. 2529–2548, 2023

    2023

  2. [2]

    Advancing embodied intelligence in robotic-assisted endovascular procedures: A systematic review of ai solutions,

    T. Yao, B. Lu, M. Kowarschik, Y . Yuan, H. Zhao, S. Ourselin, K. Althoefer, J. Ge, and P. Qi, “Advancing embodied intelligence in robotic-assisted endovascular procedures: A systematic review of ai solutions,”arXiv preprint arXiv:2504.15327, 2025. 7

    arXiv 2025

  3. [3]

    Modeling non-linear effects in a 4-dof robotic transcatheter delivery system,

    N. U. Nayar and J. P. Desai, “Modeling non-linear effects in a 4-dof robotic transcatheter delivery system,”IEEE Robotics and Automation Letters, 2024

    work page 2024

  4. [4]

    Robotic transcatheter tricuspid valve replacement with hybrid enhanced intelligence: a new paradigm and first-in-vivo study

    S. Wang, H. Lin, Y . Xie, Z. Wang, D. Chen, L. Tan, X. Hou, C. Chen, X.-H. Zhou, S. Lin,et al., “Robotic transcatheter tricuspid valve replacement with hybrid enhanced intelligence: a new paradigm and first-in-vivo study,”arXiv preprint arXiv:2411.12478, 2024

    work page Pith review arXiv 2024

  5. [5]

    Magnetic soft microfiberbots for robotic embolization,

    X. Liu, L. Wang, Y . Xiang, F. Liao, N. Li, J. Li, J. Wang, Q. Wu, C. Zhou, Y . Yang,et al., “Magnetic soft microfiberbots for robotic embolization,”Science Robotics, vol. 9, no. 87, p. eadh2479, 2024

    work page 2024

  6. [6]

    Autonomous guidewire navigation in dynamic environments,

    V . Scarponi, F. Lecomte, M. Duprez, F. Nageotte, and S. Cotin, “Autonomous guidewire navigation in dynamic environments,” in2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp. 6931–6938, IEEE, 2024

    work page 2024

  7. [7]

    Vascularpilot3d: Toward a 3d fully autonomous navigation for endovascular robotics,

    J. Song, K. Yang, H. Chen, J. Liu, Y . Gu, Q. Hui, Y . Huang, M. Li, Z. Zhang, T. Cao,et al., “Vascularpilot3d: Toward a 3d fully autonomous navigation for endovascular robotics,”arXiv preprint arXiv:2405.09375, 2024

    arXiv 2024

  8. [8]

    Sim2real learning with domain ran- domization for autonomous guidewire navigation in robotic-assisted endovascular procedures,

    T. Yao, H. Wang, B. Lu, J. Ge, Z. Pei, M. Kowarschik, L. Sun, L. Seneviratne, and P. Qi, “Sim2real learning with domain ran- domization for autonomous guidewire navigation in robotic-assisted endovascular procedures,”IEEE Transactions on Automation Science and Engineering, 2025

    work page 2025

Show all 29 references
  1. [9]

    Splineformer: An explainable transformer-based approach for autonomous endovascular navigation,

    T. Jianu, S. Doust, M. Li, B. Huang, T. Do, H. Nguyen, K. Bates, T. D. Ta, S. Fichera, P. Berthet-Rayne,et al., “Splineformer: An explainable transformer-based approach for autonomous endovascular navigation,” arXiv preprint arXiv:2501.04515, 2025

    2025 arXiv

  2. [10]

    Model-guided navigation of magnetic soft guidewire for safe endovascular surgery,

    J. Li, H. Chen, and L. Wang, “Model-guided navigation of magnetic soft guidewire for safe endovascular surgery,”Journal of the Mechan- ics and Physics of Solids, p. 105731, 2024

    2024

  3. [11]

    Sim4endor: A reinforcement learning centered simulation platform for task automation of endovas- cular robotics,

    T. Yao, M. Ban, B. Lu, Z. Pei, and P. Qi, “Sim4endor: A reinforcement learning centered simulation platform for task automation of endovas- cular robotics,”arXiv preprint arXiv:2504.05330, 2025

    2025 arXiv

  4. [12]

    Shape sensing techniques for continuum robots in minimally invasive surgery: A survey,

    C. Shi, X. Luo, P. Qi, T. Li, S. Song, Z. Najdovski, T. Fukuda, and H. Ren, “Shape sensing techniques for continuum robots in minimally invasive surgery: A survey,”IEEE Transactions on Biomedical Engi- neering, vol. 64, no. 8, pp. 1665–1678, 2016

    2016

  5. [13]

    Shape reconstruction of soft continuum robots via the fusion of local strains and global poses,

    X. An, Y . Cui, X. Dong, Y . Wang, B. Du, X.-J. Liu, and H. Zhao, “Shape reconstruction of soft continuum robots via the fusion of local strains and global poses,”Cell Reports Physical Science, vol. 5, no. 10, 2024

    2024

  6. [14]

    Autonomous intelligent navigation for flexible endoscopy using monocular depth guidance and 3-d shape planning,

    Y . Lu, R. Wei, B. Li, W. Chen, J. Zhou, Q. Dou, D. Sun, and Y .-h. Liu, “Autonomous intelligent navigation for flexible endoscopy using monocular depth guidance and 3-d shape planning,” in2023 IEEE international conference on robotics and automation (ICRA), pp. 1–7, IEEE, 2023

    2023

  7. [15]

    Fusion of biplane fluoroscopy with fiber bragg grating for 3d catheter shape reconstruction,

    M. Ourak, S. De Buck, X. T. Ha, O. Al-Ahmad, K. Bamps, J. Ector, and E. Vander Poorten, “Fusion of biplane fluoroscopy with fiber bragg grating for 3d catheter shape reconstruction,”IEEE Robotics and Automation Letters, vol. 6, no. 4, pp. 6505–6512, 2021

    2021

  8. [16]

    Desectbot: Design and validation of a novel two-segment decoupled continuum robotic system for endoscopic submucosal dissection,

    W. Liu, Y . Shao, Y . Zhang, Z. Chen, D. Wu, Y . Chen, C. Stefanini, L. Li, and P. Qi, “Desectbot: Design and validation of a novel two-segment decoupled continuum robotic system for endoscopic submucosal dissection,” in2024 IEEE/RSJ International Conference on Intelligent Rob...

    2024

  9. [17]

    Shape-guided configuration-aware learning for endoscopic- image-based pose estimation of flexible robotic instruments,

    Y . Ma, K. Chen, H.-S. Tong, R. Wei, Y .-L. Ng, K.-W. Kwok, and Q. Dou, “Shape-guided configuration-aware learning for endoscopic- image-based pose estimation of flexible robotic instruments,” inEu- ropean Conference on Computer Vision, pp. 259–276, Springer, 2024

    2024

  10. [18]

    A learning-driven framework with spatial optimization for surgical suture thread reconstruction and autonomous grasping under multiple topologies and environmental noises,

    B. Lu, W. Chen, Y .-M. Jin, D. Zhang, Q. Dou, H. K. Chu, P.- A. Heng, and Y .-H. Liu, “A learning-driven framework with spatial optimization for surgical suture thread reconstruction and autonomous grasping under multiple topologies and environmental noises,” in2020 IEEE/RSJ I...

    2020

  11. [19]

    Toward image-guided automated suture grasping under complex environments: A learning-enabled and optimization-based holistic framework,

    B. Lu, B. Li, W. Chen, Y . Jin, Z. Zhao, Q. Dou, P.-A. Heng, and Y . Liu, “Toward image-guided automated suture grasping under complex environments: A learning-enabled and optimization-based holistic framework,”IEEE Transactions on Automation Science and Engineering, vol. 19, ...

    2021

  12. [20]

    Suture thread spline reconstruction from endoscopic images for robotic surgery with reliability-driven keypoint detection,

    N. Joglekar, F. Liu, R. Orosco, and M. Yip, “Suture thread spline reconstruction from endoscopic images for robotic surgery with reliability-driven keypoint detection,” in2023 IEEE International Conference on Robotics and Automation (ICRA), pp. 4747–4753, IEEE, 2023

    2023

  13. [21]

    Enhancing percutaneous coronary intervention with heuristic path planning and deep-learning-based vascular segmentation,

    T. Yao, C. Wang, X. Wang, X. Li, Z. Jiang, and P. Qi, “Enhancing percutaneous coronary intervention with heuristic path planning and deep-learning-based vascular segmentation,”Computers in Biology and Medicine, vol. 166, p. 107540, 2023

    2023

  14. [22]

    3d guidewire shape reconstruction from monoplane fluoroscopic images,

    T. Jianu, B. Huang, P. Berthet-Rayne, S. Fichera, and A. Nguyen, “3d guidewire shape reconstruction from monoplane fluoroscopic images,” inInternational Conference on Robot Intelligence Technology and Applications, pp. 84–94, Springer, 2023

    2023

  15. [23]

    Aiareseg: Catheter detection and segmentation in interventional ultrasound using transformers,

    A. Ranne, Y . Velikova, N. Navab, and F. R. y Baena, “Aiareseg: Catheter detection and segmentation in interventional ultrasound using transformers,” in2024 IEEE International Conference on Robotics and Automation (ICRA), pp. 8187–8194, IEEE, 2024

    2024

  16. [24]

    Morphoskel3d: Morphological skeletonization of 3d point clouds for informed sampling in object classification and retrieval,

    P. Onghena, S. Velasco-Forero, and B. Marcotegui, “Morphoskel3d: Morphological skeletonization of 3d point clouds for informed sampling in object classification and retrieval,”arXiv preprint arXiv:2501.12974, 2025

    2025

  17. [25]

    Iterative pnp and its application in 3d-2d vascular image registration for robot navigation,

    J. Song, K. Yang, Z. Zhang, M. Li, T. Cao, and M. Ghaffari, “Iterative pnp and its application in 3d-2d vascular image registration for robot navigation,” in2024 IEEE International Conference on Robotics and Automation (ICRA), pp. 17560–17566, IEEE, 2024

    2024

  18. [26]

    Constraint-aware learning for fractional flow reserve pullback curve estimation from invasive coronary imaging,

    D. Zhang, X. Liu, A. Wang, H. Zhang, G. Yang, H. Zhang, and Z. Gao, “Constraint-aware learning for fractional flow reserve pullback curve estimation from invasive coronary imaging,”IEEE Transactions on Medical Imaging, 2024

    2024

  19. [27]

    2d/3d registration with a statistical deformation model prior using deep learning,

    J. Van Houtte, X. Gao, J. Sijbers, and G. Zheng, “2d/3d registration with a statistical deformation model prior using deep learning,” in 2021 IEEE EMBS international conference on biomedical and health informatics (BHI), pp. 1–4, IEEE, 2021

    2021

  20. [28]

    A reinforce- ment learning approach for real-time articulated surgical instrument 3d pose reconstruction,

    K. Fan, Z. Chen, Q. Liu, G. Ferrigno, and E. De Momi, “A reinforce- ment learning approach for real-time articulated surgical instrument 3d pose reconstruction,”IEEE Transactions on Medical Robotics and Bionics, 2024

    2024

  21. [29]

    Efficient data-driven scene simulation using robotic surgery videos via physics-embedded 3d gaussians,

    Z. Yang, K. Chen, Y . Long, and Q. Dou, “Efficient data-driven scene simulation using robotic surgery videos via physics-embedded 3d gaussians,”arXiv preprint arXiv:2405.00956, 2024. 8

    2024 arXiv

Pith tools

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