Towards cognitive supervision in robot-assisted surgery

Loris Fichera, Leonardo S. Mattos and Darwin G. Caldwell, IIT, Italy

From the initial futuristic vision of autonomous surgical robots (ASRs), able to operate even without the supervision of a skilled surgeon, researchers are now moving, due to acceptance, economical and complexity issues, towards the more realistic concept of surgical robots as intelligent tools at the direct command of the surgeon. In this perspective, artificial intelligence (AI) could help in making surgical robots more intelligent than they are today, e.g. giving them the ability to understand what the surgeon is doing, thus, making them more effective in their assistive task. In this paper, we first review some interesting examples of how machine learning (ML), a branch of AI, has been successfully applied to the field of surgical robotics. Finally, we will give an overview of how AI is likely to play an important role in the development of future surgical platforms, where the use of untangible technologies, such as lasers and ultrasound, will require cognitive abilities that go beyond human ones.