Role of robotics as a key platform for digital advancements in surgery Free

Throughout the last two centuries, surgery has encountered four major revolutions. The first revolution started with anatomical mastery, asepsis and anaesthesia and thereafter, progression towards complex multi-organ surgeries, organ transplants and prosthetics occurred. In recent decades, video-assisted minimally invasive surgery and robotic surgery have reduced invasiveness and arguably increased the precision of surgical procedures. Currently we stand at the edge of the era of digital surgery, which involves the incorporation of cutting-edge technologies like extended reality (incorporating virtual reality, augmented reality and mixed reality), telepresence and artificial intelligence (AI) to enhance surgical training, precision, efficiency and outcome¹. This fourth revolution is based on availability of computational resources, democratization of information and connectivity with high-bandwidth internet in combination with increased access to data centres offering data storage, centralization and processing. A digitized world with nearly unrestrained data access and computing power enables new research avenues aiming to optimize surgical performance and enhance patient care and recovery. The concept of Surgical Data Science uses all available local patient data from electronic health records, intra- and perioperative sensor data combined with globally available data and evidence to improve treatment efficacy and safety.

Robotic surgery should be considered an essential stepping stone and a central platform for the seamless clinical integration and culmination of the fourth surgical revolution. When compared to classical laparoscopic surgery, robotic systems hold several advantages for advancing digital surgery. First, without the need for additional sensors, robotic systems can provide information on instruments location, orientation and movement (so-called ‘kinematics’), as well as system-event data—clutching, energy device usage, stapling and camera movement. Second, the camera point-of-view in robotic surgery is more stable as it is no longer handheld. This creates the potential for overlay systems and 3D applications, which can be linked to camera shifts. Third, the video stream is accessible in 3D, enabling more accurate depth estimation and informing the user of tool–tissue interaction. Fourth, most tools and instruments are integrated in one platform, which facilitates precise and objective capture of all surgical actions for the purposes of training and technique optimization. As such, through its integrated data and sensor capabilities, robotic surgery acts as both as an efficient surgical end effector as well as a continuous data source for digital surgery. The advent of novel competitors in the robotic surgery market is expected to further accelerate the use of surgical robots as data reservoirs and platforms for digital applications. Novel robotic surgery contenders are increasingly making their intraoperative system data available to users and the wider surgical community, whereas before this was only possible under strict research collaborations.

Robotic surgery also has the potential to improve patient outcomes by means of improved surgical training. Boal and colleagues² recently evaluated the use of objective tools and AI in robotic surgery technical skills assessment in a systematic review. They found great potential for AI in supporting surgical training in combination with robotic surgery, yet also described the need for further validation and standardization in this field. General surgical performance rating tools like GEARS-Likert-scale assessments were identified as a widely accepted tool for robotic surgical training outcome evaluations. Meanwhile, studies with procedure- and task-specific tools and metrics have recently demonstrated better psychometric properties to assess surgical technical skill than GEARS-like assessments³. Proficiency-based progression is considered the current gold standard in surgical training but requires frequent objective assessments. Automated surgical skill assessment with AI thus holds tremendous potential in this field. As a promising concept, procedure-specific training in combination with procedure-specific reproducible wet lab or 3D organ models are expected to facilitate automated AI-assisted scoring through replication of an identical surgical scene^4,5. This underlines the need for increased efforts in the field of objective surgical skill assessment to inform the development of basic and specialty training curricula. Robotic platforms will facilitate both acquisition of sensor data and integration of meaningful digitized feedback to support trainees and training programmes.

Apart from sensor data and training, robotic surgery systems are ideal user interfaces for the integration of intraoperative assistance and support. The endoscopic and kinematic information presented to the surgeon pass through a computer, allowing for manipulation and augmentation of the information stream at different levels. The endoscopic stream can be augmented using preoperative 3D models, intraoperative imaging such as ultrasound or advanced tissue information such as fluorescence or hyperspectral imaging, with all of the most recent technological advancements utilizing AI to push the boundaries of what is possible⁶. The kinematic or auditive streams can likewise be augmented or altered, as such providing haptic or auditive surgical feedback or even partial automation. The feasibility of telesurgery has already been demonstrated⁷ and advances in robust and fast internet connectivity have enabled telesurgery, thus realizing the original intent of robotic surgery. Telestration is available in robotic systems via dual-console ghost tools and touch screens and telesurgical assistance can be integrated for remote support and proctoring.

In conclusion, a robotic platform's dual ability to simultaneously act as a sensor device and computer interface, hold tremendous potential for advancing intraoperative surgical support. This duality will simultaneously facilitate the development as well as the integration of real-time AI-based surgical systems. However, before fully harnessing the potential of surgical AI in robotics, it is imperative to address several fundamental requirements—for example, to establish comprehensive surgical databases through systematic availability and collation of kinematic readouts from all robotic surgery vendors. Care should be taken to not solely focus on robotic surgery for advancements in the digital surgery era, as the global uptake of robotic surgery is inequitable and would simply exaggerate the disparity in global surgical digitalization. Additionally, fostering agreement within the surgical community regarding data structuring and annotation for intelligent analysis remains a challenge⁸. Even basic steps such as annotating instruments⁹ and defining specific starting and ending points of surgical phases¹⁰ have not been agreed for computer vision applications. Dedicated procedure-specific training pathways are expected to help to systematically break down procedures into the clinically most important steps and phases. Finally, while the ability to intelligently analyse surgical video streams is a significant step forward, its value remains limited if we cannot establish a direct correlation with patient factors, pathology and outcomes. As such, future collaborative outcome-driven research remains central to integrating digital surgical solutions and facilitating this fourth revolution in surgical practice.

Author contributions

Pieter De Backer (Conceptualization, Writing—original draft), and Felix Nickel (Conceptualization, Supervision, Writing—original draft)

Funding

The authors have no funding to declare.

Disclosure

The authors declare no conflict of interest. This study was not based on a previous communication to a society or meeting.

Data availability

This study did not include any original data but was based on publicly available literature.

References