Key points are not available for this paper at this time.
Recent advances in digital pathology and artificial intelligence (AI) are transforming our ability to diagnose myeloid neoplasms, including acute myeloid leukemia (AML), myelodysplastic syndromes (MDS), and myeloproliferative neoplasms (MPN). Modern AI systems now achieve expert-level performance across peripheral blood smear assessment, bone marrow aspirate and biopsy interpretation, flow cytometry, cytogenetics, and next-generation sequencing; with many platforms surpassing 90% accuracy and enabling detection of subtle morphologic and immunophenotypic signatures beyond human perception. Moreover, AI tools increasingly integrate multimodal inputs to support classification, risk stratification, and treatment predictions, often outperforming conventional methods. These systems augment pathologists by streamlining workflows, enhancing consistency, reducing interobserver variability, and accelerating turnaround times, paving the way for advanced precision diagnostics in hematopathology. However, challenges remain in clinical adoption, including cost, security, regulatory oversight, and validation across diverse populations. Despite these hurdles, early regulatory successes such as digital peripheral blood smear morphology analyzers illustrate the feasibility of real-world implementation. Broader adoption will however require coordinated efforts among developers, laboratories, pathologists, and regulatory agencies to ensure safety, transparency, and interoperability.
Patel et al. (Fri,) studied this question.