Artificial intelligence (AI) and robotics can benefit surgery together and individually.
Not all robotic surgical systems are AI systems. Robots involve hardware and feedback. (Remember: sense-think-act.)
AI-enhanced robotics has the potential to enable automation in surgery at varying levels of autonomy, including assistance (currently in use), supervised/conditional autonomy (in prototype use), and full autonomy (in development).
Future applications of AI could include virtual assistants and mentors, improved situational awareness, and an interface to a collective surgical consciousness.
The field of surgery can gain a great deal from the convergence of artificial intelligence (AI) and robotic technologies. These related but distinct topics—as elaborated in this chapter—stand to benefit from substantial synergy. At this time, the goal of combining robotics with AI for many is to break down surgical tasks into subtasks, which can be recognized through computer vision and machine learning and then be automated using surgical robots. This is to increase procedural efficiency, accuracy, and safety while decreasing surgeon focus on automatable/rote tasks and overall fatigue, which could enable the surgeon to have increased vigilance and focus on the more complex aspects of the procedure. Future possibilities range from virtual robotic assistants that physically move in response to natural language commands or automatic visual interpretation of surgical scenes, virtual mentors that can make remote physical adjustments in the operative field that are available on demand, and autonomous robotic executions of surgical tasks or whole procedures. In this chapter, we provide an introduction to robotics and how robotics interacts with AI. This is followed by a review of the literature on robotics, AI, and the convergence of robotics and AI in the operating room, including additional possibilities for robotics and AI in the future.
Robots Are Cyber-Physical Systems That Sense, Think, and Act
Cyber-physical systems are systems that have both computational (software) and physical (hardware) components.1 AI (see Chapter 1) refers to a set of techniques within the computational aspects of automated decision making and thus fits in the “cyber” part of cyber-physical systems. Automation systems do exist that are purely algorithmic/computational, without physical components, and such systems are not robotic systems.2 Similarly, automation systems also exist that are purely mechanical/physical, without computation, and such systems are not AI systems. Such systems are also not robotic, as robots by definition involve some aspect of computation, at least by the definition chosen in this chapter of robots as cyber-physical systems that sense, think, and act with feedback control. Multiple definitions exist for robots and for AI and have been debated and discussed for decades. Furthermore, not all robotic systems are AI systems, as the computational aspects of a robotic system may not be particularly intelligent. Robotic systems with AI have both physical components for sensing, computation, and actuation, as well as algorithmic ...