ConVibNet: Needle Detection during Continuous Insertion via Frequency-Inspired Features

Dianye Huang; Jiamei Guo; Maria Neiiendam; Nassir Navab; Zhehao Duan; Zhongliang Jiang

arxiv: 2603.01147 · v1 · submitted 2026-03-01 · 💻 cs.CV

ConVibNet: Needle Detection during Continuous Insertion via Frequency-Inspired Features

Jiamei Guo , Zhehao Duan , Maria Neiiendam , Dianye Huang , Nassir Navab , Zhongliang Jiang This is my paper

Pith reviewed 2026-05-15 17:54 UTC · model grok-4.3

classification 💻 cs.CV

keywords needle detectionultrasound imagingtemporal modelingdeep learningreal-time trackingmedical interventioncontinuous insertion

0 comments

The pith

ConVibNet tracks needle tips and angles in real-time ultrasound by modeling motion correlations between frames.

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

The paper presents ConVibNet as an extension of VibNet that processes sequences of ultrasound frames to locate the needle tip and shaft angle during continuous insertion. It adds an intersection-and-difference loss to make the model explicitly sensitive to how the tip moves from one frame to the next. This approach is motivated by the fact that needles are often poorly visible in single ultrasound images due to artifacts and low contrast, so temporal information becomes essential for reliable tracking. On the authors' dataset the method reaches 2.80 mm tip error and 1.69 deg angle error while running at real-time speed, a 0.75 mm gain over the strongest baseline.

Core claim

ConVibNet extends VibNet to continuous needle insertion by leveraging temporal dependencies across successive ultrasound frames and introducing a novel intersection-and-difference loss that explicitly models motion correlations between consecutive frames, yielding a tip localization error of 2.80 mm and an angle error of 1.69 deg on the curated evaluation set.

What carries the argument

intersection-and-difference loss, which computes explicit motion correlations between consecutive frames to reinforce temporal awareness of needle-tip position

If this is right

Enables continuous estimation of both needle tip position and shaft angle in dynamic insertion scenarios.
Preserves real-time inference speed suitable for clinical deployment.
Supports integration into autonomous or robotic needle-insertion systems.
Improves robustness to intermittent visibility, occlusions, and low-contrast conditions typical in ultrasound.

Where Pith is reading between the lines

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

The same temporal-loss pattern could be tested on other thin moving structures such as catheters or guidewires in ultrasound.
Combining the loss with additional sensor data such as force or optical tracking might further reduce error without slowing inference.
If the motion-correlation term generalizes, it could shorten the training data needed for new needle-insertion tasks.

Load-bearing premise

The curated dataset and intersection-and-difference loss together produce reliable gains on unseen clinical ultrasound sequences that contain varying artifacts and insertion speeds.

What would settle it

Running ConVibNet and the best baseline on a fresh collection of ultrasound videos recorded at different insertion speeds and with different artifact patterns, then checking whether the 0.75 mm tip-error advantage disappears.

read the original abstract

Purpose: Ultrasound-guided needle interventions are widely used in clinical practice, but their success critically depends on accurate needle placement, which is frequently hindered by the poor and intermittent visibility of needles in ultrasound images. Existing approaches remain limited by artifacts, occlusions, and low contrast, and often fail to support real-time continuous insertion. To overcome these challenges, this study introduces a robust real-time framework for continuous needle detection. Methods: We present ConVibNet, an extension of VibNet for detecting needles with significantly reduced visibility, addressing real-time, continuous needle tracking during insertion. ConVibNet leverages temporal dependencies across successive ultrasound frames to enable continuous estimation of both needle tip position and shaft angle in dynamic scenarios. To strengthen temporal awareness of needle-tip motion, we introduce a novel intersection-and-difference loss that explicitly leverages motion correlations across consecutive frames. In addition, we curated a dedicated dataset for model development and evaluation. Results: The performance of the proposed ConVibNet model was evaluated on our dataset, demonstrating superior accuracy compared to the baseline VibNet and UNet-LSTM models. Specifically, ConVibNet achieved a tip error of 2.80+-2.42 mm and an angle error of 1.69+-2.00 deg. These results represent a 0.75 mm improvement in tip localization accuracy over the best-performing baseline, while preserving real-time inference capability. Conclusion: ConVibNet advances real-time needle detection in ultrasound-guided interventions by integrating temporal correlation modeling with a novel intersection-and-difference loss, thereby improving accuracy and robustness and demonstrating high potential for integration into autonomous insertion systems.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

ConVibNet adds temporal modeling and a new intersection-and-difference loss to VibNet for continuous ultrasound needle tracking, claiming a 0.75 mm tip error improvement on their dataset while staying real-time.

read the letter

The main point is that this paper extends VibNet with temporal dependencies across frames and introduces an intersection-and-difference loss to better handle continuous needle insertion in ultrasound, where visibility is poor. They report a tip error of 2.80 mm and angle error of 1.69 deg on their curated dataset, beating the original VibNet and UNet-LSTM by 0.75 mm in tip accuracy without losing real-time speed. The loss is meant to capture motion correlations between consecutive frames, which aligns with the dynamic nature of insertion procedures. They also assembled their own dataset, which addresses a common bottleneck in this area. These elements give the work a practical focus on a clinical task with real patient stakes. The framing around artifacts, occlusions, and the need for continuous estimation is straightforward and ties back to existing frequency-inspired methods. The concrete error metrics and the claim of preserved inference speed are useful for anyone thinking about deployment. The soft spots center on missing experimental transparency. The abstract and summary give no dataset size, patient or insertion counts, speed ranges, artifact variety, train-test split details, or ablation results isolating the new loss. This makes it difficult to assess whether the reported gains are robust or tied to their specific curation. Generalization to unseen clinical sequences with varying conditions remains an open question, as the stress-test note flags. If the full paper supplies those checks and shows the loss actually drives the improvement, the contribution strengthens; otherwise the central claim stays provisional. This is for researchers in medical computer vision and interventional imaging who work on ultrasound tracking or autonomous systems. A reader could extract the loss formulation or dataset curation tactic for their own setups. It deserves peer review because the task is relevant, the baseline extension is clear, and the numbers are specific enough to evaluate, even with the gaps that referees would likely probe.

Referee Report

3 major / 2 minor

Summary. The paper introduces ConVibNet, an extension of VibNet for real-time continuous needle detection and tracking in ultrasound images during insertion. It incorporates temporal dependencies across frames and proposes a novel intersection-and-difference loss to improve tip position and shaft angle estimation under poor visibility. A dedicated dataset is curated for training and evaluation. On this dataset, ConVibNet reports a tip error of 2.80 ± 2.42 mm and angle error of 1.69 ± 2.00 deg, representing a 0.75 mm improvement in tip accuracy over the best baseline while maintaining real-time inference.

Significance. If the reported gains are shown to be robust, this work could meaningfully advance real-time needle localization in ultrasound-guided interventions by addressing intermittent visibility through explicit temporal correlation modeling. The emphasis on preserving real-time capability is a practical strength for potential integration into clinical or autonomous systems.

major comments (3)

[Results] Results section: the reported tip error of 2.80 ± 2.42 mm and 0.75 mm improvement over baselines are presented without any dataset size, number of insertions/patients, train-test split description, or statistical significance testing, leaving the central accuracy claim unverifiable.
[Methods] Methods section (loss function): the intersection-and-difference loss is introduced to leverage motion correlations, yet no ablation isolating its effect from other architectural changes or the curated dataset is provided, so its specific contribution to the error reduction cannot be assessed.
[Dataset] Dataset section: the curation of a dedicated dataset is highlighted for robustness to artifacts and varying insertion speeds, but no statistics on scale, speed range, artifact types, or patient variability are supplied, undermining claims of generalization to unseen clinical sequences.

minor comments (2)

[Abstract] Abstract: the performance claims could briefly note the scale of the evaluation dataset to give readers immediate context for the reported errors.
[Methods] Notation: the description of the intersection-and-difference loss would benefit from an explicit equation or pseudocode to clarify how it differs from standard temporal losses.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive comments, which highlight important aspects of verifiability and completeness. We address each major comment below and will incorporate the necessary revisions to strengthen the manuscript.

read point-by-point responses

Referee: [Results] Results section: the reported tip error of 2.80 ± 2.42 mm and 0.75 mm improvement over baselines are presented without any dataset size, number of insertions/patients, train-test split description, or statistical significance testing, leaving the central accuracy claim unverifiable.

Authors: We agree that the Results section requires additional details for full verifiability. In the revised manuscript, we will add the total dataset size (number of frames and insertions), number of patients, a description of the train-test split (e.g., splits performed at the insertion level to prevent leakage), and statistical significance testing (e.g., paired t-tests or Wilcoxon tests) to substantiate the reported 0.75 mm improvement. revision: yes
Referee: [Methods] Methods section (loss function): the intersection-and-difference loss is introduced to leverage motion correlations, yet no ablation isolating its effect from other architectural changes or the curated dataset is provided, so its specific contribution to the error reduction cannot be assessed.

Authors: We acknowledge that an ablation study is needed to isolate the loss function's contribution. We will add an ablation analysis in the revised manuscript comparing the full ConVibNet against variants without the intersection-and-difference loss (while retaining the temporal architecture and dataset), thereby quantifying its specific impact on tip and angle accuracy. revision: yes
Referee: [Dataset] Dataset section: the curation of a dedicated dataset is highlighted for robustness to artifacts and varying insertion speeds, but no statistics on scale, speed range, artifact types, or patient variability are supplied, undermining claims of generalization to unseen clinical sequences.

Authors: We will expand the Dataset section to include the requested statistics: overall scale (total sequences and frames), insertion speed ranges, specific artifact types (e.g., shadowing, reverberation), and patient variability (number of subjects, anatomical variations). This will better support the generalization claims. revision: yes

Circularity Check

0 steps flagged

No significant circularity in derivation chain

full rationale

The paper introduces ConVibNet as an extension of VibNet with a novel intersection-and-difference loss and a curated dataset. Reported performance (tip error 2.80±2.42 mm, angle error 1.69±2.00 deg, 0.75 mm improvement over baselines) is evaluated on held-out portions of the dataset against external baselines (VibNet, UNet-LSTM). No equations, derivations, or self-citations reduce the claimed results to quantities defined by the same fitted parameters or inputs by construction. The evaluation uses independent held-out data and external comparisons, leaving the central claims with independent content.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

The central claim rests on standard supervised deep-learning assumptions plus the effectiveness of the newly introduced loss; no explicit free parameters or invented physical entities are stated in the abstract.

axioms (1)

domain assumption Neural networks trained with the proposed loss will generalize from the curated dataset to new clinical ultrasound sequences
Implicit in any claim that the reported errors will hold in practice.

invented entities (1)

intersection-and-difference loss no independent evidence
purpose: To explicitly leverage motion correlations across consecutive frames for tip-motion awareness
New loss term introduced in the methods to strengthen temporal modeling

pith-pipeline@v0.9.0 · 5613 in / 1218 out tokens · 56108 ms · 2026-05-15T17:54:42.878845+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

ConVibNet leverages temporal dependencies across successive ultrasound frames... novel intersection-and-difference loss... L = Lf(t) + Lf(t+Δ) + α Linter + β Ldiff
IndisputableMonolith/Foundation/RealityFromDistinction.lean reality_from_one_distinction unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

frequency-domain feature extraction... STFT spectrograms of needle tip/shaft vs tissue

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

21 extracted references · 21 canonical work pages

[1]

Ultrasonics69, 38–46 (2016)

Kuang, Y., Hilgers, A., Sadiq, M., Cochran, S., Corner, G., Huang, Z.: Modelling 11 and characterisation of a ultrasound-actuated needle for improved visibility in ultrasound-guided regional anaesthesia and tissue biopsy. Ultrasonics69, 38–46 (2016)

work page 2016
[2]

IEEE Transactions on Medical Imaging44(6), 2696–2708 (2025)

Huang, D., Li, C., Karlas, A., Chu, X., Samuel Au, K.W., Navab, N., Jiang, Z.: Vibnet: Vibration-boosted needle detection in ultrasound images. IEEE Transactions on Medical Imaging44(6), 2696–2708 (2025)

work page 2025
[3]

IEEE Robotics and Automation Letters (2025)

Zhang, Y., Lei, L., Yan, W., Zhang, T., Tang, R.S.-Y., Cheng, S.S.: Mambax- ctrack: Mamba-based tracker with ssm cross-correlation and motion prompt for ultrasound needle tracking. IEEE Robotics and Automation Letters (2025)

work page 2025
[4]

arXiv preprint arXiv:2312.01239 (2023)

Goel, R., Morales, C., Singh, M., Dubrawski, A., Galeotti, J., Choset, H.: Motion informed needle segmentation in ultrasound images. arXiv preprint arXiv:2312.01239 (2023)

work page arXiv 2023
[5]

IEEE Transactions on Control Systems Technology25(3), 966–978 (2016)

Mathiassen, K., Dall’Alba, D., Muradore, R., Fiorini, P., Elle, O.J.: Robust real- time needle tracking in 2-d ultrasound images using statistical filtering. IEEE Transactions on Control Systems Technology25(3), 966–978 (2016)

work page 2016
[6]

In: 2015 International Conference on Advanced Robotics (ICAR), pp

Kaya, M., Senel, E., Ahmad, A., Orhan, O., Bebek, O.: Real-time needle tip localization in 2d ultrasound images for robotic biopsies. In: 2015 International Conference on Advanced Robotics (ICAR), pp. 47–52 (2015). IEEE

work page 2015
[7]

Medical image analysis89, 102878 (2023)

Jiang, Z., Salcudean, S.E., Navab, N.: Robotic ultrasound imaging: State-of-the- art and future perspectives. Medical image analysis89, 102878 (2023)

work page 2023
[8]

Annual Review of Control, Robotics, and Autonomous Systems7(2024)

Bi, Y., Jiang, Z., Duelmer, F., Huang, D., Navab, N.: Machine learning in robotic ultrasound imaging: Challenges and perspectives. Annual Review of Control, Robotics, and Autonomous Systems7(2024)

work page 2024
[9]

The International Journal of Robotics Research43(7), 981–1002 (2024)

Jiang, Z., Bi, Y., Zhou, M., Hu, Y., Burke, M., Navab, N.: Intelligent robotic sonographer: Mutual information-based disentangled reward learning from few demonstrations. The International Journal of Robotics Research43(7), 981–1002 (2024)

work page 2024
[10]

International journal of computer assisted radiology and surgery13(5), 647–657 (2018)

Mwikirize, C., Nosher, J.L., Hacihaliloglu, I.: Convolution neural networks for real-time needle detection and localization in 2d ultrasound. International journal of computer assisted radiology and surgery13(5), 647–657 (2018)

work page 2018
[11]

Ultrasonic Imaging43(5), 262–272 (2021)

Wijata, A., Andrzejewski, J., Pyci´ nski, B.: An automatic biopsy needle detection and segmentation on ultrasound images using a convolutional neural network. Ultrasonic Imaging43(5), 262–272 (2021)

work page 2021
[12]

In: 2023 IEEE/RSJ Inter- national Conference on Intelligent Robots and Systems (IROS), pp

Huang, D., Bi, Y., Navab, N., Jiang, Z.: Motion magnification in robotic sonog- raphy: Enabling pulsation-aware artery segmentation. In: 2023 IEEE/RSJ Inter- national Conference on Intelligent Robots and Systems (IROS), pp. 6565–6570 12 (2023). IEEE

work page 2023
[13]

In: 2021 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), pp

Belikova, K., Zailer, A., Tekucheva, S.V., Ermoljev, S.N., Dylov, D.V.: Deep learning for spatio-temporal localization of temporomandibular joint in ultra- sound videos. In: 2021 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), pp. 1257–1261 (2021). IEEE

work page 2021
[14]

Journal of Ultrasound in Medicine41(2), 311–325 (2022)

Hovgesen, C.H., Wilhjelm, J.E., Vilmann, P., Kalaitzakis, E.: Echogenic sur- face enhancements for improving needle visualization in ultrasound: A prisma systematic review. Journal of Ultrasound in Medicine41(2), 311–325 (2022)

work page 2022
[15]

Brachytherapy22(6), 761–768 (2023)

Dupere, J.M., Brost, E.E., Uthamaraj, S., Lee, C.U., Urban, M.W., Stish, B.J., Deufel, C.L.: A new way to visualize prostate brachytherapy needles using ultra- sound color doppler and needle surface modifications. Brachytherapy22(6), 761–768 (2023)

work page 2023
[16]

International journal of computer assisted radiology and surgery11(6), 1183–1192 (2016)

Beigi, P., Rohling, R., Salcudean, S.E., Ng, G.C.: Spectral analysis of the tremor motion for needle detection in curvilinear ultrasound via spatiotemporal linear sampling. International journal of computer assisted radiology and surgery11(6), 1183–1192 (2016)

work page 2016
[17]

Ultrasonics78, 18–22 (2017)

Beigi, P., Rohling, R., Salcudean, T., Lessoway, V.A., Ng, G.C.: Detection of an invisible needle in ultrasound using a probabilistic svm and time-domain features. Ultrasonics78, 18–22 (2017)

work page 2017
[18]

In: 2017 IEEE International Conference on Computer Vision (ICCV), pp

Lin, T.-Y., Goyal, P., Girshick, R., He, K., Doll´ ar, P.: Focal loss for dense object detection. In: 2017 IEEE International Conference on Computer Vision (ICCV), pp. 2999–3007 (2017)

work page 2017
[19]

IEEE Transactions on Instrumentation and Measurement (2024)

Jiang, Z., Li, X., Chu, X., Karlas, A., Bi, Y., Cheng, Y., Au, K.S., Navab, N.: Needle segmentation using gan: Restoring thin instrument visibility in robotic ultrasound. IEEE Transactions on Instrumentation and Measurement (2024)

work page 2024
[20]

IEEE Transactions on Automation Science and Engineering22, 381–392 (2024)

Huang, D., Yang, C., Zhou, M., Karlas, A., Navab, N., Jiang, Z.: Robot-assisted deep venous thrombosis ultrasound examination using virtual fixture. IEEE Transactions on Automation Science and Engineering22, 381–392 (2024)

work page 2024
[21]

Medical Engineering and Physics29(4), 413–431 (2007) 13

Abolhassani, N., Patel, R., Moallem, M.: Needle insertion into soft tissue: A survey. Medical Engineering and Physics29(4), 413–431 (2007) 13

work page 2007

[1] [1]

Ultrasonics69, 38–46 (2016)

Kuang, Y., Hilgers, A., Sadiq, M., Cochran, S., Corner, G., Huang, Z.: Modelling 11 and characterisation of a ultrasound-actuated needle for improved visibility in ultrasound-guided regional anaesthesia and tissue biopsy. Ultrasonics69, 38–46 (2016)

work page 2016

[2] [2]

IEEE Transactions on Medical Imaging44(6), 2696–2708 (2025)

Huang, D., Li, C., Karlas, A., Chu, X., Samuel Au, K.W., Navab, N., Jiang, Z.: Vibnet: Vibration-boosted needle detection in ultrasound images. IEEE Transactions on Medical Imaging44(6), 2696–2708 (2025)

work page 2025

[3] [3]

IEEE Robotics and Automation Letters (2025)

Zhang, Y., Lei, L., Yan, W., Zhang, T., Tang, R.S.-Y., Cheng, S.S.: Mambax- ctrack: Mamba-based tracker with ssm cross-correlation and motion prompt for ultrasound needle tracking. IEEE Robotics and Automation Letters (2025)

work page 2025

[4] [4]

arXiv preprint arXiv:2312.01239 (2023)

Goel, R., Morales, C., Singh, M., Dubrawski, A., Galeotti, J., Choset, H.: Motion informed needle segmentation in ultrasound images. arXiv preprint arXiv:2312.01239 (2023)

work page arXiv 2023

[5] [5]

IEEE Transactions on Control Systems Technology25(3), 966–978 (2016)

Mathiassen, K., Dall’Alba, D., Muradore, R., Fiorini, P., Elle, O.J.: Robust real- time needle tracking in 2-d ultrasound images using statistical filtering. IEEE Transactions on Control Systems Technology25(3), 966–978 (2016)

work page 2016

[6] [6]

In: 2015 International Conference on Advanced Robotics (ICAR), pp

Kaya, M., Senel, E., Ahmad, A., Orhan, O., Bebek, O.: Real-time needle tip localization in 2d ultrasound images for robotic biopsies. In: 2015 International Conference on Advanced Robotics (ICAR), pp. 47–52 (2015). IEEE

work page 2015

[7] [7]

Medical image analysis89, 102878 (2023)

Jiang, Z., Salcudean, S.E., Navab, N.: Robotic ultrasound imaging: State-of-the- art and future perspectives. Medical image analysis89, 102878 (2023)

work page 2023

[8] [8]

Annual Review of Control, Robotics, and Autonomous Systems7(2024)

Bi, Y., Jiang, Z., Duelmer, F., Huang, D., Navab, N.: Machine learning in robotic ultrasound imaging: Challenges and perspectives. Annual Review of Control, Robotics, and Autonomous Systems7(2024)

work page 2024

[9] [9]

The International Journal of Robotics Research43(7), 981–1002 (2024)

Jiang, Z., Bi, Y., Zhou, M., Hu, Y., Burke, M., Navab, N.: Intelligent robotic sonographer: Mutual information-based disentangled reward learning from few demonstrations. The International Journal of Robotics Research43(7), 981–1002 (2024)

work page 2024

[10] [10]

International journal of computer assisted radiology and surgery13(5), 647–657 (2018)

Mwikirize, C., Nosher, J.L., Hacihaliloglu, I.: Convolution neural networks for real-time needle detection and localization in 2d ultrasound. International journal of computer assisted radiology and surgery13(5), 647–657 (2018)

work page 2018

[11] [11]

Ultrasonic Imaging43(5), 262–272 (2021)

Wijata, A., Andrzejewski, J., Pyci´ nski, B.: An automatic biopsy needle detection and segmentation on ultrasound images using a convolutional neural network. Ultrasonic Imaging43(5), 262–272 (2021)

work page 2021

[12] [12]

In: 2023 IEEE/RSJ Inter- national Conference on Intelligent Robots and Systems (IROS), pp

Huang, D., Bi, Y., Navab, N., Jiang, Z.: Motion magnification in robotic sonog- raphy: Enabling pulsation-aware artery segmentation. In: 2023 IEEE/RSJ Inter- national Conference on Intelligent Robots and Systems (IROS), pp. 6565–6570 12 (2023). IEEE

work page 2023

[13] [13]

In: 2021 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), pp

Belikova, K., Zailer, A., Tekucheva, S.V., Ermoljev, S.N., Dylov, D.V.: Deep learning for spatio-temporal localization of temporomandibular joint in ultra- sound videos. In: 2021 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), pp. 1257–1261 (2021). IEEE

work page 2021

[14] [14]

Journal of Ultrasound in Medicine41(2), 311–325 (2022)

Hovgesen, C.H., Wilhjelm, J.E., Vilmann, P., Kalaitzakis, E.: Echogenic sur- face enhancements for improving needle visualization in ultrasound: A prisma systematic review. Journal of Ultrasound in Medicine41(2), 311–325 (2022)

work page 2022

[15] [15]

Brachytherapy22(6), 761–768 (2023)

Dupere, J.M., Brost, E.E., Uthamaraj, S., Lee, C.U., Urban, M.W., Stish, B.J., Deufel, C.L.: A new way to visualize prostate brachytherapy needles using ultra- sound color doppler and needle surface modifications. Brachytherapy22(6), 761–768 (2023)

work page 2023

[16] [16]

International journal of computer assisted radiology and surgery11(6), 1183–1192 (2016)

Beigi, P., Rohling, R., Salcudean, S.E., Ng, G.C.: Spectral analysis of the tremor motion for needle detection in curvilinear ultrasound via spatiotemporal linear sampling. International journal of computer assisted radiology and surgery11(6), 1183–1192 (2016)

work page 2016

[17] [17]

Ultrasonics78, 18–22 (2017)

Beigi, P., Rohling, R., Salcudean, T., Lessoway, V.A., Ng, G.C.: Detection of an invisible needle in ultrasound using a probabilistic svm and time-domain features. Ultrasonics78, 18–22 (2017)

work page 2017

[18] [18]

In: 2017 IEEE International Conference on Computer Vision (ICCV), pp

Lin, T.-Y., Goyal, P., Girshick, R., He, K., Doll´ ar, P.: Focal loss for dense object detection. In: 2017 IEEE International Conference on Computer Vision (ICCV), pp. 2999–3007 (2017)

work page 2017

[19] [19]

IEEE Transactions on Instrumentation and Measurement (2024)

Jiang, Z., Li, X., Chu, X., Karlas, A., Bi, Y., Cheng, Y., Au, K.S., Navab, N.: Needle segmentation using gan: Restoring thin instrument visibility in robotic ultrasound. IEEE Transactions on Instrumentation and Measurement (2024)

work page 2024

[20] [20]

IEEE Transactions on Automation Science and Engineering22, 381–392 (2024)

Huang, D., Yang, C., Zhou, M., Karlas, A., Navab, N., Jiang, Z.: Robot-assisted deep venous thrombosis ultrasound examination using virtual fixture. IEEE Transactions on Automation Science and Engineering22, 381–392 (2024)

work page 2024

[21] [21]

Medical Engineering and Physics29(4), 413–431 (2007) 13

Abolhassani, N., Patel, R., Moallem, M.: Needle insertion into soft tissue: A survey. Medical Engineering and Physics29(4), 413–431 (2007) 13

work page 2007