S441 Clinical Utility and Application of Artificial Intelligence (AI) Algorithms in the Accurate Diagnosis of Appendicitis—A Meta-Analysis

Introduction: Appendicitis is a relatively common acute abdominal condition, requiring prompt and accurate diagnosis to prevent complications. Traditional diagnostic methods often lead to misdiagnosis and unnecessary surgeries. Artificial intelligence (AI) algorithms have emerged as promising tools for enhancing diagnostic accuracy. By analyzing vast amounts of clinical data, imaging studies in particular, AI can identify subtle patterns and indicators of appendicitis. Its utility mainly lies in reducing diagnostic errors and optimizing the improving patient care in the domain of appendicitis. Methods: The review conducted follows the PRISMA guidelines and major medical databases, which include PUBMED, Google Scholar and Science-Direct, were extensively searched using a comprehensive search term to identify and retrieve available articles. The articles that assessed diagnostic potential, i.e., accuracy and area under the curve (AUC), were included in the final analysis. The data was analyzed using the Meta, Metadata and the Metafor packages of R Studio. The pooled accuracy and AUC were assessed in the study to estimate the diagnostic potential of AI-assisted imaging in the diagnosis of appendicitis. The random effects model via the linear (mixed-effects) model framework was considered for statistical analysis. Results: The study included a total of 19 studies with a total of 7,767 images that assessed the presence of appendicitis. The pooled accuracy for the diagnosis of appendicitis was assessed to be 0.89 (0.84; 0.93, 95% confidence interval [CI], P < 0.0001). The pooled AUC emerged to be 0.97 (0.95; 0.99, 95% CI, P= 0.7585). Conclusion: Through this meta-analysis, the Artificial Intelligence algorithms have been analyzed and evaluated to be clinically effective in the detection of appendicitis, as evident by their high accuracy.