Artificial Intelligence Applications in Clinical Laboratory Diagnostics: a Systematic Review of Diagnostic Accuracy, Workflow Efficiency, and Clinical Utility

Abstract

Artificial intelligence (AI) is increasingly integrated into clinical laboratory diagnostics, offering advanced capabilities for data interpretation and clinical decision support. However, existing evidence remains fragmented across diagnostic performance, operational efficiency, and real-world clinical impact.
This systematic review aimed to evaluate the applications of AI in clinical laboratory diagnostics, with a focus on diagnostic accuracy, workflow efficiency, and clinical utility.
This review was conducted in accordance with PRISMA guidelines. A comprehensive search of PubMed, Scopus, Web of Science, and the Cochrane Library was performed for studies published within the last 13 years. Eligible studies included diagnostic accuracy studies, observational studies, and clinical trials evaluating AI applications in laboratory diagnostics. Data extraction and quality assessment were conducted using standardized tools, and findings were synthesized narratively.
A total of 65 studies were included from 450 identified records. AI models demonstrated consistently high diagnostic performance across multiple clinical domains, frequently achieving area under the curve (AUC) values above 0.85. Models integrating multimodal data showed enhanced robustness compared to single-modality approaches. AI applications improved laboratory workflow by automating data interpretation, reducing turnaround times, and optimizing resource utilization. In addition, AI demonstrated strong clinical utility in early disease detection, risk stratification, and personalized medicine. However, limitations such as lack of external validation, dataset heterogeneity, and limited real-world implementation were commonly reported.
AI has significant potential to transform clinical laboratory diagnostics by enhancing accuracy, efficiency, and clinical decision-making. Future research should prioritize multicenter validation, standardized evaluation frameworks, and real-world implementation to ensure safe, equitable, and effective integration into healthcare systems.