Clinical Applications of Artificial Intelligence in Modern Hematology: Diagnostic and Prognostic Perspectives

Artificial intelligence (AI) has recently emerged as an important tool supporting diagnostics and clinical decision-making across multiple fields of medicine, including hematology. Rapid advances in machine learning algorithms and computational data analysis have enabled the processing and interpretation of complex biomedical datasets derived from laboratory diagnostics, medical imaging, and molecular studies [1,2]. These developments are particularly relevant in hematology, where the interpretation of large volumes of morphological, immunophenotypic, and genetic data plays a crucial role in disease diagnosis and monitoring. In hematological practice, artificial intelligence has demonstrated promising applications in the automated analysis of peripheral blood smears and bone marrow aspirates, as well as in the interpretation of flow cytometry data and molecular diagnostic results [3,4]. Several studies have shown that AI-based systems can achieve high levels of sensitivity and specificity in the detection of abnormal cellular populations, including leukemic blasts, potentially improving diagnostic accuracy and reducing the time required for laboratory evaluation [5]. Despite these promising developments, the clinical implementation of artificial intelligence in hematology remains associated with several challenges, including the need for large, high-quality training datasets, external validation across different patient populations, and integration with existing laboratory information systems [6]. The aim of this article is to review the current state of knowledge regarding the use of artificial intelligence in hematology, with particular emphasis on its applications in laboratory diagnostics, microscopic image analysis, and prognostic modeling in hematological diseases.