Artificial intelligence in hematology: current trends and application areas

Abstract

This study aims to map the current research trends, application areas, and thematic clusters of artificial intelligence (AI) technologies in hematology. A literature search was conducted in the Web of Science on June 1, 2025, covering the years 1980–2025. A total of 376 original research articles were analyzed using bibliometric techniques, including trend keyword and factor analyses, via the bibliometrix package in R Studio. The USA (n = 111), China (79), the United Kingdom (22), and Germany (21) were the most productive countries, with Harvard University as the leading institution. Keywords revealed research concentration in thrombosis, venous thromboembolism, risk assessment, hematopoietic stem cell transplantation, atrial fibrillation, anticoagulants, morphological analysis, blood management, sepsis, and acute myeloid leukemia. Recent years have shown rising interest in transcriptome, prediction modeling, hematopoietic stem cell transplantation, blood management, thrombosis, venous thromboembolism cancer and COVID-19 related complications. Factor analysis grouped the literature into five clusters: AI and core hematologic diseases, clinical quality improvement, risk prediction using health data, intensive care, and cardiovascular applications. AI demonstrates substantial contributions to diagnostic accuracy, prognostication, and personalized care in hematology. The findings highlight AI’s growing potential in enhancing clinical decision-making and improving patient outcomes through data-driven and genomics-based innovations.