Application of Artificial Intelligence in Inborn Errors of Immunity Identification and Management: Past, Present, and Future—A Systematic Review

Background: Inborn errors of immunity (IEI) are mainly genetically driven disorders that affect immune function and present with highly heterogeneous clinical manifestations, ranging from severe combined immunodeficiency (SCID) to adult-onset immune dysregulatory diseases. This clinical heterogeneity, coupled with limited awareness and the absence of a universal diagnostic test, makes early and accurate diagnosis challenging. Although genetic testing methods such as whole-exome and genome sequencing have improved detection, they are often expensive, complex, and require functional validation. Recently, artificial intelligence (AI) tools have emerged as promising for enhancing diagnostic accuracy and clinical decision-making for IEI. Methods: We conducted a systematic review of four major databases (PubMed, Scopus, Web of Science, and Embase) to identify peer-reviewed English-published studies focusing on the application of AI techniques in the diagnosis and treatment of IEI across pediatric and adult populations. Twenty-three retrospective/prospective studies and clinical trials were included. Results: AI methodologies demonstrated high diagnostic accuracy, improved detection of pathogenic mutations, and enhanced prediction of clinical outcomes. AI tools effectively integrated and analyzed electronic health records (EHRs), clinical, immunological, and genetic data, thereby accelerating the diagnostic process and supporting personalized treatment strategies. Conclusions: AI technologies show significant promise in the early detection and management of IEI by reducing diagnostic delays and healthcare costs. While offering substantial benefits, limitations such as data bias and methodological inconsistencies among studies must be addressed to ensure broader clinical applicability.