State-of-the-Art: A Taxonomy of AI-Assisted Robotics for Medical Therapies and Applications

This paper presents a survey on the development and major advances of artificial intelligence assisted robotics for therapeutic tasks by concentrating on the current challenges emerging from the clinical application process and the research efforts mitigating the problems. In this survey, we search Nature, Science, Cell and other websites with high influence by using keywords (i.e., artificial assisted medical robots), categorized research works over the past three decades based on therapeutic applications, and discussed the latest development and bottleneck problems of each subtopic finally. Specifically, we first present a chronology of the artificial intelligence assisted techniques developed for medical therapeutic tasks over the past three decades and then classify them according to the principles of the algorithm and its corresponding type of medical therapeutic tasks. The artificial intelligence technologies in the chronology evolve from classic machine learning statistical methods of the early nineties to data driven deep learning methods. Then a taxonomy of the artificial intelligence technologies assisted therapeutic tasks in the past three decades is described according to the therapeutic task types and hot topics of the knotty problems. One prosperous trend has been abstracted from the interpretation of our taxonomy and the most highly cited research papers using certain search criteria with Nature and Cell databases, which undergoes revolutionary development of artificial intelligence and closer integration with clinical therapeutic tasks. The trend is unprecedent and more comprehensive Human-Robot Interaction, which benefits sophisticated telesurgery and microsurgery by being capable of facilitating Surgeons with higher imaging accuracy and human-like tactile sensation. Our survey discusses the current grand challenges and future trends of artificial intelligence assisted therapeutic tasks for the convenience of clinical research and applications. We hope this survey would help bridging the gap between entrepreneurial translation and research.