Preoperative diagnosis of periprosthetic joint infection in patients undergoing total hip and knee arthroplasty revision: development and validation of a machine learning algorithm

This project aims to develop a machine learning algorithm to improve the preoperative diagnosis of periprosthetic joint infections (PJIs) in total hip and knee arthroplasty revisions. PJIs, affecting ~2% of cases, are challenging to diagnose due to limitations in existing criteria and false negatives from bacterial biofilms. By analyzing clinical and radiological data, the algorithm seeks to enhance diagnostic accuracy, enabling timely and precise treatments. The two-phase study involves algorithm development using retrospective data (2015–2020) and validation on 350 patients, comparing its performance against EBJIS criteria. This innovation promises to refine antibiotic therapies, optimize surgical strategies, and support personalized care, advancing AI’s role in orthopedic decision-making.