Artificial Intelligence in Revision Total Knee Arthroplasty

This chapter reviews recent developments on the role of artificial intelligence in revision total knee arthroplasty (rTKA). Machine learning (ML) can detect prosthetic loosening from radiographs in first-time rTKA with an accuracy of about 86%. Artificial intelligence algorithms predict prolonged length of stay following rTKA. ML models predict periprosthetic joint infection (PJI) following rTKA for aseptic failure, recurrent infection after rTKA for PJI, venous thromboembolism and major bleeding events following rTKA, and 30-day mortality, cardiovascular adverse events, and respiratory adverse events after aseptic rTKA.