Attention: Restrictions on use of AUA, AUAER, and UCF content in third party applications, including artificial intelligence technologies, such as large language models and generative AI.
You are prohibited from using or uploading content you accessed through this website into external applications, bots, software, or websites, including those using artificial intelligence technologies and infrastructure, including deep learning, machine learning and large language models and generative AI.

ROBOTICS Robotic Surgery and Artificial Intelligence: A Synergistic Nexus

By: Mitchell G. Goldenberg, MBBS, PhD, Keck School of Medicine, University of Southern California, Los Angeles, Artificial Intelligence Center, USC Institute of Urology, University of Southern California, Los Angeles; Severin Rodler, MD, Keck School of Medicine, University of Southern California, Los Angeles, Artificial Intelligence Center, USC Institute of Urology, University of Southern California, Los Angeles; Giovanni Cacciamani, MSc, MD, FEBU, Keck School of Medicine, University of Southern California, Los Angeles, Artificial Intelligence Center, USC Institute of Urology, University of Southern California, Los Angeles | Posted on: 01 Mar 2024

Artificial intelligence (AI) continues its march into the mainstream of clinical urology, with new applications regularly appearing in different facets of patient care. As robotic-assisted surgery (RAS) has become ubiquitous in our specialty, our horizons have broadened to see this platform as a “jumping-off point” for future disruptive and digital technology in surgical care. The exponential growth of AI capabilities seen in other high-reliability industries has only been minimally integrated into daily patient care, its pace appropriately slowed by concerns around transparency and accountability. As with autonomous vehicles, the immediate impact on human safety in this space cannot be overstated, and any steps toward bringing AI into the operating room must be approached with an abundance of caution. Despite these concerns, a monumental shift in how we provide surgical care to patients with urological conditions has appeared on the horizon, and AI’s incursion into the world of RAS is now inevitable.

RAS is the perfect medium to facilitate AI’s introduction to the operating room (Figure). Originally designed as a means of providing surgeons with the ability to carry out minimally invasive surgery with advanced dexterity, control, and visualization, the robotic platform’s embrace of innovation is entwined in its very DNA. The interface between human and machine provides the ideal setting for the utilization of machine learning as a tool to enhance surgical education, patient safety, and operative efficiency. Early successful integration into these fields has been widely published and celebrated in the academic community. Evidence demonstrates that AI can accurately classify surgeons by their level of skill, matching expert surgeon evaluations with a high level of accuracy.1 This has important implications for important educational initiatives such as competency-based medical education, an approach to residency training that demands high volume, frequent assessments of performance.2 The use of AI to reliably provide an appraisal of a trainee’s surgical skill using only audiovisual inputs will provide a level of objectivity and scalability that has been identified as a limitation of reliance on human-expert evaluation alone. Beyond identification of trainees in need of remediation, AI also has shown promise as a vehicle for providing objective, performance-based feedback to trainees in the simulation lab,3 an undoubtedly underutilized educational resource at most institutions. While these efforts originally used machine learning trained on kinematic data derived from the robotic console or physical instrument trackers,4 there has been a shift toward computer-vision AI approaches in this area that rely solely on video data taken from the robotic endoscope to make predictions. This unencumbered approach to AI-augmented skills assessment appears to be generalizable across different surgical techniques and procedures, and the near-complete lack of reliance on human data labeling will facilitate the dissemination of this technology.

Figure. Artificial intelligence applications for robotic surgery. AR indicates augmented reality; iAE, intraoperative adverse events.

Perhaps the most compelling use of AI in this space is its potential to improve and even standardize patient outcomes; specifically, the real-time recognition of threats to patient safety intraoperatively. Recognition of surgeon errors is becoming possible in real-time, providing surgical teams with the ability to correct deviations in procedural steps that may go otherwise overlooked, but also mitigate the potential harm that results from intraoperative adverse events through immediate recognition.5 Perhaps even more remarkable is the demonstration that algorithms using computer vision can accurately identify surgical phases using video data alone.6 When intraoperative adverse events mitigation and surgical phase detection are aligned, it seems we are close to the advent of real-time, predictive analytics in the operating room—that is, the ability for AI to guide surgeons through an operation, a true “second set of eyes” that can help surgeons with complex intraoperative decision making.7

While still in its infancy, automation of robotic tasks using AI is being explored as both a means of reducing human error and increasing surgical efficiency. While the idea of a surgical robot carrying out multiple steps of an operation has not yet been realized, there are examples in the literature of robotics and AI being married to allow for the execution of simple tasks such as suturing and knot tying.8 It is important to remember that while many would imagine a scene from science fiction, with an autonomous machine carrying out unsupervised advanced procedures with impossible speed, there are more subtle surgical actions whose automation is a more realistic endeavor in the near future. These include optimization of instrument position and camera view to allow for more precise and efficient surgeon movements, integration of patient imaging with augmented reality, and tissue interaction sensing (ie, tension, torque, etc).8 The most immediate uses of AI in RAS do not need to involve surrendering control of the operation to a machine, but rather using this technology to improve our ability to carry these operations out consistently and safely.

The future is bright for RAS, and the permeation of AI into our daily lives in and out of the operating room will only grow. Robotic surgery is now becoming the new standard for many surgical procedures,9 and the shift from “robotic-assisted” to “AI-assisted” surgery is just a matter of time (and clear regulations!!). It is imperative that as clinicians we are the stewards of this technology, questioning not only the accuracy and precision of these algorithms, but the ethical aspects of introducing what is essentially another decision-maker into the sacred physician-patient relationship. There are obvious barriers that need to be overcome on the path to implementing AI into routine surgical care on a wide scale, least of which are the medicolegal ramifications of any reliance on these algorithms to determine how we perform in the operating room or make determinations related to the competency of a fellow surgeon or trainee.10 It is our hope that our early forays into this space are thought provoking and hypothesis generating, and as this technology becomes more tangible for all clinicians, we can continue to explore how AI can improve the well-being of both our patients and ourselves.

  1. Khalid S, Goldenberg M, Grantcharov T, Taati B, Rudzicz F. Evaluation of deep learning models for identifying surgical actions and measuring performance. JAMA Netw Open. 2020;3(3):e201664.
  2. Holmboe ES, Sherbino J, Long DM, Swing SR, Frank JR. The role of assessment in competency-based medical education. Med Teach. 2010;32(8):676-682.
  3. Ma R, Lee RS, Nguyen JH, et al. Tailored feedback based on clinically relevant performance metrics expedites the acquisition of robotic suturing skills—an unblinded pilot randomized controlled trial. J Urol. 2022;208(2):414-424.
  4. Hung AJ, Chen J, Gill IS. Automated performance metrics and machine learning algorithms to measure surgeon performance and anticipate clinical outcomes in robotic surgery. JAMA Surg. 2018;153(8):770-771.
  5. Eppler M, Sayegh A, Maas M, et al. Automated capture of intraoperative adverse events using artificial intelligence: a systematic review and meta-analysis. J Clin Med. 2023;12(4):1687.
  6. Garrow CR, Kowalewski K-F, Li L, et al. Machine learning for surgical phase recognition. Ann Surg. 2021;273(4):684-693.
  7. Colborn K, Brat G, Callcut R. Predictive analytics and artificial intelligence in surgery—opportunities and risks. JAMA Surg. 2023;158(4):337.
  8. Attanasio A, Scaglioni B, De Momi E, Fiorini P, Valdastri P. Autonomy in surgical robotics. Annu Rev Control Robot Auton Syst. 2021;4(1):651-679.
  9. Gill I, Cacciamani G. LBA3 the changing face of urologic oncologic surgery from 2000-2018 (63 141 patients)-impact of robotics. J Urol. 2018;199(4S):e577-e578.
  10. Jovanovic I. AI in endoscopy and medicolegal issues: the computer is guilty in case of missed cancer?. Endosc Int Open. 2020;08(10):E1385-E1386.