Current perspectives and challenges of using artificial intelligence in immunodeficiencies

Abstract

The rapid growth of artificial intelligence (AI) in health care is promising for screening and early diagnosis in settings that heavily rely on professional expertise, such as rare diseases like inborn errors of immunity (IEI). However, the development of AI algorithms for IEI and other rare diseases faces important challenges such as dataset sizes, availability and harmonization. Similarly, the implementation of AI-based strategies for screening and diagnosis of IEI in real-world scenarios is hampered by multiple factors including stakeholders' acceptance, ethical and legal constraints, and technologic barriers. Consequently, while the body of literature on AI-based solutions for early diagnosis of IEI continues to expand, clinical utility and widespread implementation remain limited. In this review, we provide an up-to-date comprehensive review of current applications and challenges facing AI use for IEI diagnosis and care.