An Exhaustive Analysis of AI Applications in Robotic Surgery Focus on the Art

Abstract

While there is a lot of written about AI's great potential, no studies have looked at how well it improves patient safety during robot-assisted surgery (RAS). In accordance with the PRISMA 2020 declaration, a literature search was carried out using PubMed, Web of Science, Scopus, and IEEExplore. Articles published in English between January 1, 2016, and December 31, 2020 that were peer-reviewed were considered. For quality assessment, Amstar 2 was utilized. The Newcastle Ottawa Quality assessment instrument was used to evaluate the risk of bias. The SPIDER tool was used to graphically exhibit the study data in tables. The search parameters were satisfied by 35 articles, which comprised 3436 patients in the study. Among the chosen studies are those pertaining to training (n = 3), tissue retraction (n = 1), urology (n = 12), gynecology (n = 1), and other specialties (n = 1). The detection precision of surgical instruments varies rising from 76.0 percent to 90.6 percent. After a robot-assisted radical prostatectomy (RARP), the average absolute error for predicting urine continence was between 85.9 and 134.7 days. A forecast accuracy of 88.5% was achieved for the duration of stay after RARP. During robot-assisted partial nephrectomy (RAPN), the next surgical job was accurately recognized 75.7% of the time. Overall, the quality of the studies that were considered was poor. Due to the restricted size of the datasets, the conclusions are rather constrained. Because methods and datasets are different, it was difficult to compare research that dealt with the same subject. The important tasks of RAS procedures impact patient outcome, but there is no indication that AI can detect them at this time. Immediate action is required to validate AI systems through external means and conduct studies on massive datasets. Surgeons should also be able to understand and use the data to communicate with patients in a way that is understandable to the average person.