Fairness in AI: How Can We Avoid Bias and Disparities in Orthopedic Applications of Artificial Intelligence?

Abstract

Recent advances in artificial intelligence have the potential to transform the field of orthopedics. As well as the opportunities there are numerous challenges associated with applying AI to clinical decision-making, one such example being algorithmic fairness. In this article we introduce the concepts of bias and fairness in machine learning from an orthopedics perspective, covering concepts, examples, possible approaches and implications on the community. We hope that by working to embed these concepts and associated best practice into health data-product development workflows, we can help to promote fair and effective use of these powerful tools for all patients.