**5.1 Surgical education**

Perhaps, the most significant benefit of AR in surgery is in medical education. Head-mounted devices used in AR have already proven useful in various aspects of medical education, including anatomy and surgery [83]. In the near future, AR may allow surgeons to practice various procedures anywhere in a low-stakes environment and decrease cognitive effort, allowing for a more sustained practice [84]. AR may eventually be used within the operating room as a teaching tool, allowing surgeons to manipulate personalized models of the patient's organs based on some of the techniques described previously. Thus, AR may become a valuable supplemental tool to train future surgeons and other specialists who want to practice procedures.

Machine Learning algorithms may play other essential roles in surgical education. Aspiring surgeons start their training with varying degrees of motor skill and learning abilities, with the use of ML algorithms in the future, students may be able to be classified based on generated learning curves. Gao et al. were able to analyze the proficiency of students performing various surgical tasks using an algorithm to predict the number of trials needed for each student to proficiently complete the task [85]. Similar algorithms in the future may be applied to planning surgical resources for students based on the need to optimize learning for all students within a surgical program. Other ML programs may be able to provide feedback to learners about specific skills. For instance, surgical skill is an important factor in patient outcomes, directly preventing complications and indirectly in mediating other elements such as the length

*Artificial Intelligence in Surgery, Surgical Subspecialties, and Related Disciplines DOI: http://dx.doi.org/10.5772/intechopen.112691*

of surgery [86]. Thus, measuring and improving surgical skills is important in improving patient care. However, there is a lack of practical objective assessments of surgical skill and dexterity. Currently, many assessments of surgical skills are subjective in nature [87]. AI algorithms may be able to address these concerns.

Video-based learning remains a promising learning method for surgical residents [88]. However, video-based review can be limited by having to parse through long videos, especially when reviewing multiple examples. Hashimoto et al. show that it is possible to develop a computer vision model capable of accurately identifying distinct phases of a surgery [89]. This technology allows surgeons to quickly find specific stages of an operation for more efficient review, and similar AI models have been validated in other types of surgeries as well [90]. While out of the scope of these studies, these models could be supplemented with AI that directly analyzes the surgeon's skills. For instance, an algorithm could be created to rate surgical motion economy within the operation theater, and by proxy surgical skill [91]. Using videos of surgeons performing the same procedure, the algorithm may be able to provide objective feedback on the motion economy and path length compared to other surgeons in a video database. AI programs that combine surgical phase recognition and surgical skill analysis could be used to indicate certain stages of the procedure where the surgeon could improve motion economy. Surgeons, especially those in training, may not be completely aware of unnecessary movements they are making during surgery, and these algorithms could provide an objective way to compare and teach motion economy. AI algorithms may be applied to similar measures, such as fluidity of motion, force application in laparoscopic surgery, or a combination of these factors. In the future, these algorithms may provide objective insight into surgical skills and dexterity, allowing for targeted practice of specific skills (**Table 3**).




**Table 3.**

*Summary of included studies on computer vision and augmented reality (AR).*
