AI-Assisted Cholecystectomy Study

Overview

The objective of this prospective clinical pilot study is to evaluate the clinical feasibility, safety, and workflow integration of the voice-based AI-assisted robotic system for performing selected auxiliary tasks during robotic cholecystectomy.

Robotic-assisted surgery provides surgeons with enhanced dexterity, stable three-dimensional visualization, and improved ergonomics for minimally invasive procedures. However, despite these technological advances, complex robotic operations remain cognitively demanding and highly coordinated. In addition to controlling the primary operative instruments, surgeons are required to manage auxiliary tasks such as tissue retraction, maintenance of surgical exposure, and coordination of multiple robotic arms. These auxiliary tasks often require frequent switching between control modes, manual adjustments, or verbal coordination with clinical assistants, which may interrupt surgical flow and increase cognitive workload. In current clinical practice, maintaining stable and optimal exposure of the surgical field is essential for safe dissection and precise manipulation. Elective robotic or laparoscopic cholecystectomy is a standardized and well-established surgical procedure that involves repetitive and well-defined sub-tasks such as tissue retraction, exposure adjustment, and cystic duct and artery clipping. During this procedure, surgeons must repeatedly adjust retraction direction and tension in response to anatomical variation, tissue deformation, and procedural progression. These adjustments are commonly performed by manually controlling an additional robotic arm or by communicating with an assistant surgeon. Such workflows may divert attention from the primary operative task and increase the risk of inefficiency or distraction, particularly during delicate dissection near critical anatomical structures. To address the challenges associated with auxiliary task management in robotic surgery, we propose a novel voice-assisted automated "third hand" system for supervised surgical assistance. Recent advances in artificial intelligence, robotic perception, and natural language processing have enabled the development of interactive systems capable of understanding spoken commands and executing structured actions in real time. Prior work suggests that voice-based interaction can be feasibly applied in human-machine and human-robot collaboration, indicating its potential as a hands-free control interface \[1,2\]. At the same time, progress in task-specific robotic control has made it increasingly feasible to deploy limited, well-defined automation under continuous human supervision. Together, these developments create opportunities for designing intelligent assistive systems that support surgical workflow while preserving surgeon authority and responsibility. The proposed system integrates a natural language interface with task-specific robotic control to enable surgeons to verbally command and adjust a dedicated robotic arm responsible for tissue retraction, exposure maintenance, and selected repetitive actions. Through predefined motion constraints, force limits, and continuous monitoring, the system is designed to function as an assistive tool rather than an autonomous decision-making agent. The surgeon remains fully responsible for procedural planning and execution and may interrupt or override automated actions at any time. The research team has successfully undergone multiple pre-clinical trials including ex-vivo porcine, live porcine as well as cadaveric procedures, confirming the safety of the AI system. (Refer to "Prior Study" section) Elective robotic cholecystectomy provides an appropriate experimental model for evaluating this approach due to its standardized workflow and frequent requirement for dynamic exposure adjustment. By delegating selected auxiliary tasks to an AI-controlled robotic arm while maintaining continuous surgeon supervision, this study enables systematic evaluation of the feasibility, controllability, and safety of limited, task-specific automation. This experiment, therefore, proposes a clinical first-in-human pilot study to evaluate a collaborative surgical framework in which an automated third hand executes constrained sub-tasks under direct surgeon oversight, while voice commands are used to dynamically adjust and coordinate robotic behavior during surgery, without replacing surgeon judgment or decision-making

Conditions

Interventions

Eligibility

Locations (1)

Outcomes