pith. machine review for the scientific record. sign in

arxiv: 2604.11423 · v1 · submitted 2026-04-13 · 💻 cs.RO

Recognition: unknown

Dyadic Partnership(DP): A Missing Link Towards Full Autonomy in Medical Robotics

Nassir Navab , Zhongliang Jiang
Authors on Pith no claims yet

Pith reviewed 2026-05-10 15:43 UTC · model grok-4.3

classification 💻 cs.RO
keywords medical roboticshuman-robot collaborationdyadic partnershiprobotic autonomyfoundation modelsintent recognitionexplainable AI
0
0 comments X

The pith

Medical robots and clinicians can collaborate as dyadic partners to discuss decisions and gradually reach full autonomy.

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

The paper claims that decades of tele-manipulation robots have delivered dexterity and imaging gains but left all cognition and decisions to the surgeon. It introduces dyadic partnership as the missing intermediate stage in which robots and clinicians engage in intelligent, two-way interaction to agree on actions during procedures. This stage relies on generative AI for world models, multi-modal visualization, and components such as foundation models, intent recognition, co-learning, and trust-aware interfaces. A sympathetic reader would care because the proposal offers a concrete, acceptable route from current limited-intelligence systems to safer partial and full autonomy without requiring an abrupt leap.

Core claim

Dyadic Partnership is a new paradigm in which robots and clinicians engage in intelligent, expert interaction and collaboration. The Dyadic Partners discuss and agree on decisions and actions during their dynamic and interactive collaboration relying also on intuitive advanced media using generative AI, such as a world model, and advanced multi-modal visualization. This article outlines the foundational components needed to enable such systems, including foundation models for clinical intelligence, multi-modal intent recognition, co-learning frameworks, advanced visualization, and explainable, trust-aware interaction.

What carries the argument

Dyadic Partnership (DP), the two-agent collaboration framework in which robot and clinician discuss and jointly decide actions using generative-AI world models and multi-modal visualization.

If this is right

  • Robots and clinicians can dynamically discuss and jointly validate surgical decisions during live procedures.
  • Foundation models and co-learning frameworks supply the clinical intelligence currently missing from tele-manipulation systems.
  • Advanced visualization and trust-aware interfaces make the robot's reasoning legible to the clinician in real time.
  • The same architecture supports a staged rollout from partial to full autonomy across multiple surgical domains.

Where Pith is reading between the lines

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

  • Existing tele-manipulation consoles could be augmented rather than replaced, lowering the barrier to clinical adoption.
  • The approach may generalize beyond surgery to other high-stakes human-robot domains such as interventional radiology.
  • Successful dyadic systems would generate new datasets of agreed-upon actions that could accelerate training of future autonomous agents.

Load-bearing premise

The listed components can be integrated into a working dyadic system that improves outcomes over current tele-manipulation without creating new failure modes.

What would settle it

A controlled clinical comparison in which dyadic-partnership prototypes show no measurable gain in safety, task completion time, or surgeon workload, or introduce new error types such as misread intent or delayed agreement.

read the original abstract

For the past decades medical robotic solutions were mostly based on the concept of tele-manipulation. While their design was extremely intelligent, allowing for better access, improved dexterity, reduced tremor, and improved imaging, their intelligence was limited. They therefore left cognition and decision making to the surgeon. As medical robotics advances towards high-level autonomy, the scientific community needs to explore the required pathway towards partial and full autonomy. Here, we introduce the concept of Dyadic Partnership(DP), a new paradigm in which robots and clinicians engage in intelligent, expert interaction and collaboration. The Dyadic Partners would discuss and agree on decisions and actions during their dynamic and interactive collaboration relying also on intuitive advanced media using generative AI, such as a world model, and advanced multi-modal visualization. This article outlines the foundational components needed to enable such systems, including foundation models for clinical intelligence, multi-modal intent recognition, co-learning frameworks, advanced visualization, and explainable, trust-aware interaction. We further discuss key challenges such as data scarcity, lack of standardization, and ethical acceptance. Dyadic partnership is introduced and is positioned as a powerful yet achievable, acceptable milestone offering a promising pathway toward safer, more intuitive collaboration and a gradual transition to full autonomy across diverse clinical settings.

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.

Referee Report

2 major / 2 minor

Summary. The manuscript introduces Dyadic Partnership (DP) as a new paradigm for medical robotics that bridges current tele-manipulation systems and future full autonomy. In this framework, robots and clinicians engage in dynamic, intelligent collaboration on decisions and actions, supported by components including foundation models for clinical intelligence, multi-modal intent recognition, co-learning frameworks, advanced visualization with generative AI (such as world models), and explainable trust-aware interaction. The paper outlines these foundational elements, identifies challenges such as data scarcity, lack of standardization, and ethical acceptance, and positions DP as an achievable, acceptable milestone toward safer and more intuitive clinical robotics.

Significance. If the proposed integration of components can be realized without introducing unacceptable new risks, the DP concept could provide a useful organizing framework for research on human-robot collaboration in medicine, potentially accelerating progress from limited-intelligence teleoperation toward higher autonomy while maintaining clinician oversight. As a purely conceptual position piece, however, its significance hinges on future empirical validation of feasibility.

major comments (2)
  1. [Abstract and foundational components section] Abstract and the section outlining foundational components: The central claim that DP constitutes an 'achievable' and 'promising pathway' requires that the five listed components can be combined into a working system that improves outcomes over tele-manipulation without new failure modes (e.g., latency, conflicting decisions, or amplified errors). No architecture sketch, interaction diagram, data-flow analysis, or argument addressing these integration risks is supplied, leaving the achievability assertion unsupported.
  2. [Challenges discussion] Discussion of challenges: The manuscript correctly flags data scarcity, lack of standardization, and ethical acceptance as obstacles but provides no concrete strategies, references to existing mitigation approaches, or analysis of how the proposed DP components would specifically address them in safety-critical clinical settings; this weakens the practicality of the overall proposal.
minor comments (2)
  1. [Abstract] Abstract: 'Dyadic Partnership(DP)' is missing a space and should read 'Dyadic Partnership (DP)'.
  2. [Abstract and components outline] Throughout: The phrase 'world model' is introduced in the context of generative AI and advanced media but is neither defined nor accompanied by a reference, which reduces clarity for readers outside the immediate subfield.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback and for recognizing the potential of the Dyadic Partnership (DP) framework as an organizing concept for human-robot collaboration in medicine. As a conceptual position paper, our goal is to introduce the paradigm and its components rather than present an implemented system; we address the specific concerns below and outline targeted revisions.

read point-by-point responses

  1. Referee: [Abstract and foundational components section] Abstract and the section outlining foundational components: The central claim that DP constitutes an 'achievable' and 'promising pathway' requires that the five listed components can be combined into a working system that improves outcomes over tele-manipulation without new failure modes (e.g., latency, conflicting decisions, or amplified errors). No architecture sketch, interaction diagram, data-flow analysis, or argument addressing these integration risks is supplied, leaving the achievability assertion unsupported.

    Authors: We agree that the manuscript would benefit from a more explicit illustration of component integration. The paper is a position piece arguing that DP provides a safer intermediate step by retaining clinician oversight while incorporating foundation models and multi-modal interfaces; achievability is framed conceptually as building on mature tele-manipulation platforms. To strengthen this, we will add a high-level architecture diagram in the revised manuscript depicting data flows, decision hierarchies, and safeguards (e.g., latency buffering and conflict resolution via priority rules). A brief accompanying paragraph will address integration risks without claiming empirical validation. revision: partial

  2. Referee: [Challenges discussion] Discussion of challenges: The manuscript correctly flags data scarcity, lack of standardization, and ethical acceptance as obstacles but provides no concrete strategies, references to existing mitigation approaches, or analysis of how the proposed DP components would specifically address them in safety-critical clinical settings; this weakens the practicality of the overall proposal.

    Authors: We accept this critique and will expand the challenges section. The revision will incorporate references to established approaches such as federated learning and synthetic data augmentation for scarcity, alignment with emerging ISO 13482 and FDA AI/ML guidance for standardization, and the role of explainable AI in building ethical acceptance. We will also add a mapping table showing how each DP component (e.g., co-learning frameworks for data efficiency, trust-aware interaction for ethics) can incrementally mitigate these issues in clinical workflows. revision: yes

Circularity Check

0 steps flagged

No circularity: conceptual proposal with no derivations or fits

full rationale

The paper introduces the Dyadic Partnership concept as a new paradigm and outlines required components (foundation models, multi-modal intent recognition, co-learning, visualization, trust-aware interaction) plus challenges (data scarcity, standardization, ethics). No equations, quantitative predictions, fitted parameters, or self-citation chains exist. The positioning of DP as 'achievable' and a 'promising pathway' is a forward-looking claim without any reduction to prior author-defined quantities or self-referential derivations. This is a standard non-circular conceptual outline.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 1 invented entities

The central claim rests on the unproven premise that the enumerated AI and interface technologies can be combined into a reliable dyadic system. No free parameters or quantitative fits appear. The new entity is the Dyadic Partnership itself.

axioms (2)
  • domain assumption Foundation models trained on clinical data can provide reliable clinical intelligence for real-time decision support.
    Invoked when listing foundation models as a foundational component without citing performance bounds or failure cases in surgical contexts.
  • domain assumption Multi-modal intent recognition and generative visualization will enable intuitive, low-latency agreement between robot and clinician.
    Assumed in the description of advanced media and interaction without supporting evidence or pilot data.
invented entities (1)
  • Dyadic Partnership (DP) no independent evidence
    purpose: An intermediate collaboration paradigm positioned between tele-manipulation and full autonomy.
    New framing introduced in the paper to organize the listed components; no independent falsifiable prediction or external benchmark is supplied.

pith-pipeline@v0.9.0 · 5517 in / 1503 out tokens · 41132 ms · 2026-05-10T15:43:46.767586+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

44 extracted references · 4 canonical work pages · 1 internal anchor

  1. [1]

    Science robotics6(60), 8017 (2021)

    Dupont, P.E., Nelson, B.J., Goldfarb, M., Hannaford, B., Menciassi, A., O’Malley, M.K., Simaan, N., Valdastri, P., Yang, G.-Z.: A decade retrospective of medical robotics research from 2010 to 2020. Science robotics6(60), 8017 (2021)

    2010

  2. [2]

    Nature Reviews Bioengineering, 1–14 (2025)

    Ciuti, G., Webster III, R.J., Kwok, K.-W., Menciassi, A.: Robotic surgery. Nature Reviews Bioengineering, 1–14 (2025)

    2025

  3. [3]

    World Health Organization, ??? (2023)

    Organization, W.H.: Tracking Universal Health Coverage: 2023 Global Monitor- ing Report. World Health Organization, ??? (2023)

    2023

  4. [4]

    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)

    2023

  5. [5]

    Science Robotics10(104), 1874 (2025)

    Alterovitz, R., Hoelscher, J., Kuntz, A.: Medical needles in the hands of ai: Advancing toward autonomous robotic navigation. Science Robotics10(104), 1874 (2025)

    2025

  6. [6]

    Annual Review of Control, Robotics, and Autonomous Systems7

    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

  7. [7]

    Science Robotics10(104), 0684 (2025)

    Yip, M.: The robot will see you now: Foundation models are the path forward for 15 autonomous robotic surgery. Science Robotics10(104), 0684 (2025)

    2025

  8. [8]

    Science Robotics10(104), 0187 (2025)

    Schmidgall, S., Opfermann, J.D., Kim, J.W., Krieger, A.: Will your next surgeon be a robot? autonomy and ai in robotic surgery. Science Robotics10(104), 0187 (2025)

    2025

  9. [9]

    Science Robotics10(104), 8279 (2025)

    Dupont, P.E., Degirmenci, A.: The grand challenges of learning medical robot autonomy. Science Robotics10(104), 8279 (2025)

    2025

  10. [10]

    American Association for the Advancement of Science (2025)

    Dupont, P.E.: Medical robots learn to be autonomous. American Association for the Advancement of Science (2025)

    2025

  11. [11]

    American Association for the Advancement of Science (2017)

    Yang, G.-Z., Cambias, J., Cleary, K., Daimler, E., Drake, J., Dupont, P.E., Hata, N., Kazanzides, P., Martel, S., Patel, R.V., et al.: Medical robotics—Regulatory, ethical, and legal considerations for increasing levels of autonomy. American Association for the Advancement of Science (2017)

    2017

  12. [12]

    Annual Review of Control, Robotics, and Autonomous Systems 4(1), 651–679 (2021)

    Attanasio, A., Scaglioni, B., De Momi, E., Fiorini, P., Valdastri, P.: Autonomy in surgical robotics. Annual Review of Control, Robotics, and Autonomous Systems 4(1), 651–679 (2021)

    2021

  13. [13]

    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)

    2024

  14. [14]

    IEEE Transactions on Robotics23(1), 4–19 (2007)

    Li, M., Ishii, M., Taylor, R.H.: Spatial motion constraints using virtual fixtures generated by anatomy. IEEE Transactions on Robotics23(1), 4–19 (2007)

    2007

  15. [15]

    2, 2022- 06-27

    LeCun, Y.: A path towards autonomous machine intelligence version 0.9. 2, 2022- 06-27. Open Review62(1), 1–62 (2022)

    2022

  16. [16]

    World Models

    Ha, D., Schmidhuber, J.: World models. arXiv preprint arXiv:1803.101222(3) (2018)

    work page internal anchor Pith review arXiv 2018

  17. [17]

    In: 2007 6th IEEE and ACM Inter- national Symposium on Mixed and Augmented Reality, pp

    Bichlmeier, C., Wimmer, F., Heining, S.M., Navab, N.: Contextual anatomic mimesis hybrid in-situ visualization method for improving multi-sensory depth perception in medical augmented reality. In: 2007 6th IEEE and ACM Inter- national Symposium on Mixed and Augmented Reality, pp. 129–138 (2007). IEEE

    2007

  18. [18]

    In: International Conference on Medical Image Computing and Computer-Assisted Intervention, pp

    Lerotic, M., Chung, A.J., Mylonas, G., Yang, G.-Z.: Pq-space based non- photorealistic rendering for augmented reality. In: International Conference on Medical Image Computing and Computer-Assisted Intervention, pp. 102–109 (2007). Springer

    2007

  19. [19]

    IEEE Transactions on Medical Imaging43(6), 16 2229–2240 (2024)

    Wang, H., Ni, D., Wang, Y.: Recursive deformable pyramid network for unsuper- vised medical image registration. IEEE Transactions on Medical Imaging43(6), 16 2229–2240 (2024)

    2024

  20. [20]

    IEEE Transactions on Automation Science and Engineering22, 4818–4830 (2024)

    Jiang, Z., Kang, Y., Bi, Y., Li, X., Li, C., Navab, N.: Class-aware cartilage segmentation for autonomous us-ct registration in robotic intercostal ultra- sound imaging. IEEE Transactions on Automation Science and Engineering22, 4818–4830 (2024)

    2024

  21. [21]

    How far are surgeons from surgical world models? a pilot study on zero-shot surgical video generation with expert assessment.arXiv preprint arXiv:2511.01775, 2025

    Chen, Z., Xu, Q., Wu, J., Yang, B., Zhai, Y., Guo, G., Zhang, J., Ding, Y., Navab, N., Luo, J.: How far are surgeons from surgical world models? a pilot study on zero-shot surgical video generation with expert assessment. arXiv preprint arXiv:2511.01775 (2025)

    work page arXiv 2025

  22. [22]

    Medical Image Analysis103, 103571 (2025)

    Matinfar, S., Dehghani, S., Salehi, M., Sommersperger, M., Navab, N., Farid- pooya, K., Fairhurst, M., Navab, N.: From tissue to sound: A new paradigm for medical sonic interaction design. Medical Image Analysis103, 103571 (2025)

    2025

  23. [23]

    IROS2025 (2025)

    Zhang, Y., Huang, D., Navab, N., Jiang, Z.: Tactile-guided robotic ultrasound: Mapping preplanned scan paths for intercostal imaging. IROS2025 (2025)

    2025

  24. [24]

    Science370(6519), 966–970 (2020)

    Lee, S., Franklin, S., Hassani, F.A., Yokota, T., Nayeem, M.O.G., Wang, Y., Leib, R., Cheng, G., Franklin, D.W., Someya, T.: Nanomesh pressure sensor for monitoring finger manipulation without sensory interference. Science370(6519), 966–970 (2020)

    2020

  25. [25]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Bannur, S., Hyland, S., Liu, Q., Perez-Garcia, F., Ilse, M., Castro, D.C., Boecking, B., Sharma, H., Bouzid, K., Thieme, A.,et al.: Learning to exploit temporal struc- ture for biomedical vision-language processing. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 15016–15027 (2023)

    2023

  26. [26]

    In: Proceedings of the Conference on Empiri- cal Methods in Natural Language Processing

    Wang, Z., Wu, Z., Agarwal, D., Sun, J.: Medclip: Contrastive learning from unpaired medical images and text. In: Proceedings of the Conference on Empiri- cal Methods in Natural Language Processing. Conference on Empirical Methods in Natural Language Processing, vol. 2022, p. 3876 (2022)

    2022

  27. [27]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp

    Huang, S.-C., Shen, L., Lungren, M.P., Yeung, S.: Gloria: A multimodal global- local representation learning framework for label-efficient medical image recogni- tion. In: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 3942–3951 (2021)

    2021

  28. [28]

    Driess, D., Xia, F., Sajjadi, M.S., Lynch, C., Chowdhery, A., Wahid, A., Tomp- son, J., Vuong, Q., Yu, T., Huang, W., et al.: Palm-e: An embodied multimodal language model (2023)

    2023

  29. [29]

    In: Conference on Robot Learning, pp

    Zitkovich, B., Yu, T., Xu, S., Xu, P., Xiao, T., Xia, F., Wu, J., Wohlhart, P., Welker, S., Wahid, A.,et al.: Rt-2: Vision-language-action models transfer web knowledge to robotic control. In: Conference on Robot Learning, pp. 2165–2183 17 (2023). PMLR

    2023

  30. [30]

    IEEE Robotics and Automation Letters (2025)

    Bi, Y., Su, Y., Navab, N., Jiang, Z.: Gaze-guided robotic vascular ultrasound leveraging human intention estimation. IEEE Robotics and Automation Letters (2025)

    2025

  31. [31]

    Medical image analysis90, 102981 (2023)

    Men, Q., Teng, C., Drukker, L., Papageorghiou, A.T., Noble, J.A.: Gaze-probe joint guidance with multi-task learning in obstetric ultrasound scanning. Medical image analysis90, 102981 (2023)

    2023

  32. [32]

    Science381(6654), 141–146 (2023)

    Yip, M., Salcudean, S., Goldberg, K., Althoefer, K., Menciassi, A., Opfermann, J.D., Krieger, A., Swaminathan, K., Walsh, C.J., Huang, H.,et al.: Artificial intelligence meets medical robotics. Science381(6654), 141–146 (2023)

    2023

  33. [33]

    Science robotics7(62), 2908 (2022)

    Autonomous robotic laparoscopic surgery for intestinal anastomosis. Science robotics7(62), 2908 (2022)

    2022

  34. [34]

    Science Robotics10(104), 3093 (2025)

    Long, Y., Lin, A., Kwok, D.H.C., Zhang, L., Yang, Z., Shi, K., Song, L., Fu, J., Lin, H., Wei, W.,et al.: Surgical embodied intelligence for generalized task autonomy in laparoscopic robot-assisted surgery. Science Robotics10(104), 3093 (2025)

    2025

  35. [35]

    Nature630(8016), 353–359 (2024)

    Luo, S., Jiang, M., Zhang, S., Zhu, J., Yu, S., Dominguez Silva, I., Wang, T., Rouse, E., Zhou, B., Yuk, H.,et al.: Experiment-free exoskeleton assistance via learning in simulation. Nature630(8016), 353–359 (2024)

    2024

  36. [36]

    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)

    2024

  37. [37]

    Science387(6741), 1383–1390 (2025)

    Ha, K.-H., Yoo, J., Li, S., Mao, Y., Xu, S., Qi, H., Wu, H., Fan, C., Yuan, H., Kim, J.-T.,et al.: Full freedom-of-motion actuators as advanced haptic interfaces. Science387(6741), 1383–1390 (2025)

    2025

  38. [38]

    Robo-DM: Data Management For Large Robot Datasets.arXiv preprint arXiv:2505.15558, 2025

    Chen, K., Fu, L., Huang, D., Zhang, Y., Chen, L.Y., Huang, H., Hari, K., Balakr- ishna, A., Xiao, T., Sanketi, P.R., et al.: Robo-dm: Data management for large robot datasets. arXiv preprint arXiv:2505.15558 (2025)

    work page arXiv 2025

  39. [39]

    In: Proceedings of the Computer Vision and Pattern Recognition Conference, pp

    ¨Ozsoy, E., Pellegrini, C., Czempiel, T., Tristram, F., Yuan, K., Bani-Harouni, D., Eck, U., Busam, B., Keicher, M., Navab, N.: Mm-or: A large multimodal operating room dataset for semantic understanding of high-intensity surgical environments. In: Proceedings of the Computer Vision and Pattern Recognition Conference, pp. 19378–19389 (2025)

    2025

  40. [40]

    arXiv preprint arXiv:2505.24287 (2025)

    ¨Ozsoy, E., Mamur, A., Tristram, F., Pellegrini, C., Wysocki, M., Busam, B., Navab, N.: Egoexor: An ego-exo-centric operating room dataset for surgical 18 activity understanding. arXiv preprint arXiv:2505.24287 (2025)

    work page arXiv 2025

  41. [41]

    In: International Conference on Medical Image Computing and Computer-Assisted Intervention (2025)

    Song, T., Li, F., Bi, Y., Karlas, A., Yousefi, A., Branzan, D., Jiang, Z., Eck, U., Navab, N.: Intelligent virtual sonographer (ivs): Enhancing physician- robot-patient communication. In: International Conference on Medical Image Computing and Computer-Assisted Intervention (2025). Springer

    2025

  42. [42]

    IEEE Transactions on Visualization and Computer Graphics (2025)

    Song, T., Pabst, F., Eck, U., Navab, N.: Enhancing patient acceptance of robotic ultrasound through conversational virtual agent and immersive visualizations. IEEE Transactions on Visualization and Computer Graphics (2025)

    2025

  43. [43]

    Science390(6774), 710–715 (2025)

    Landers, F.C., Hertle, L., Pustovalov, V., Sivakumaran, D., Oral, C.M., Brinkmann, O., Meiners, K., Theiler, P., Gantenbein, V., Veciana, A.,et al.: Clinically ready magnetic microrobots for targeted therapies. Science390(6774), 710–715 (2025)

    2025

  44. [44]

    Nature machine intelligence2(10), 595–606 (2020) 19

    Martin, J.W., Scaglioni, B., Norton, J.C., Subramanian, V., Arezzo, A., Obstein, K.L., Valdastri, P.: Enabling the future of colonoscopy with intelligent and autonomous magnetic manipulation. Nature machine intelligence2(10), 595–606 (2020) 19

    2020