LAUSR

Since the emergence of Artificial Intelligence (AI) in various academic fields, the medical community has explored the usage of AI to improve current techniques and productivity in the domain of surgery. Specifically in the field of gastrointestinal (GI) surgery, research has focused on machine learning or the usage of convoluted neural network algorithms to recognize changes in intestinal mucosal and vascular patterns for identification and characterization of polyps or cancer. In this video, we demonstrate the clinical application of Computer Assisted Detection (CADe) colonoscopy to pick up various polyp types using GI GeniusTM (Fig. 1) Intelligent Endoscopy Module (US-DG-2000309 © 2021 Medtronic). Societies such as the American and European Society for Gastrointestinal Endoscopy (ASGE and ESGE) have acknowledged the diagnostic and therapeutic benefits of periodic colonoscopies. Therefore, colonoscopic screening remains the mainstay of polyp detection and colorectal cancer prevention. However, there are some limitations; the experience of an endoscopist is strongly correlated with his or her ability to detect polyps accurately. Being able to detect and determine the morphology of polyps are skills that require experience in patternrecognition and extensive knowledge acquisition—both of which demand countless hours of work. AI helps to level the playing field by aiding in polyp detection, particularly in detecting diminutive and flat polyps that are difficult to spot with the unaided eye. This ensures that most, if not all polyps with malignant potential are picked up if exposed to the AI software during colonoscopy, potentially reducing the risk of colon cancer development in these patients. Despite the benefits of AI-aided polyp detection, there exist some disadvantages. Implementing AI in healthcare mandates interaction between AI and technologically unfamiliar individuals. Naturally, some prejudice against the integration of AI into healthcare is to be expected, particularly due to the image of “dangerous AI” formed in mass culture. Furthermore, there exists resistance towards the implementation of AI by endoscopists as well. There is a common misconception that AI threatens job security. With AI, endoscopists may be overly reliant on the system, resulting in a deterioration of skill level. In fact, AI helps to improve the quality of healthcare by picking up where we might lack; serving as a tool to complement the valuable skills which endoscopists have spent time building, as well as providing a safety net to improve patient outcomes. Studies have shown that real-time AI-aided endoscopy improves adenoma detection rates as compared to the current method of independent identification of such polyps. Higher detection rates with the assistance of AI may also be observed in the context of more experienced endoscopists. Common practice is to send these resected polyps for histological evaluation which contributes to the increasing healthcare cost. AI-aided colonoscopy can assist in lowering costs by negating the need for unnecessary removal of low-risk polyps, as proposed in