In the ever-changing field of hematology, the integration of Artificial Intelligence (AI) signifies an unprecedented opportunity to transform patient care. From diagnosis to prognosis, AI, Machine Learning (ML), and Deep Learning (DL) models hold the potential to redefine our approach to hematologic disorders, spanning from diagnosis to prognosis DL methods achieve accuracies comparable to those of experts, providing faster and more precise sample processing, at a scale and pace that  supercede human capabilities.  ML techniques are poised to support clinicians in data analysis, to enhance the objectivity and accuracy of their assessments. At the same time, the knowledge embedded in these systems will assist less experienced doctors in their decision-making processes.  Undoubtedly, the future of clinical diagnosis and treatment in hematology involves the integration of AI-based hematopathology diagnostics software into routine practice.

Although hematology diagnostics primarily rely on assessing phenotypic characteristics, diagnostic uncertainty is common, and the quality of the results relies heavily on the operator’s expertise. To limit the reliance on expert knowledge and improve the consistency of data interpretation, integrating automated processes offer significant benefits. These processes aim to produce standardized, structured data, facilitating diverse projects aimed at creating dependable and consistent AI models.

In the last five years, there has been rapid progress in AI technologies, resulting in a variety of specialized AI applications applicable throughout hematology patient care, spanning from analyzing peripheral blood differentials to gene profiling. When working alongside human physicians, these AI models can enhance hematological diagnostics beyond what either could achieve independently. On one hand, incorporating ML methods will support clinicians in analyzing and interpreting data, enhancing objectivity and precision in their assessments. On the other hand, the accumulated and the integrated knowledge of these systems will assist less experienced doctors navigate their decision-making process.

Several studies have demonstrated the capabilities of AI-based models in automatically differentiating cells, detecting malignant cell populations, and supporting chromosome banding analysis, to support not only diagnosis but patient prognosis. . However, the effectiveness of these tools hinges on their appropriate application and accurate interpretation of results.

One noteworthy application is the use of DL models for automated blood cell morphology analysis,which has shown promising results in reducing interobserver variability and enhancing diagnostic accuracy. . ML algorithms have been instrumental in predicting  thrombosis risk in myeloproliferative neoplasms (MPNs) patients and identifying new leukemia subtypes based on genomic data. These models have also been applied to analyze extensive datasets of hematology images, including peripheral blood smears (PBSs), and lymph node biopsies. The result? AI-aided diagnosis systems that assist hematologists and pathologists in making more accurate diagnoses.

As AI-based technologies rapidly evolve, ensuring high-quality standards, evidence-based applications, and comprehensive validation are crucial steps to responsibly integrate ML models into clinical practice. While AI-based systems will not replace clinicians, they offer incremental benefits by assisting in decision-making processes. These systems have the potential to unburden  clinicians from tedious tasks, allowing them to focus on more complex cases and spend quality time with their patients.

In essence, AI in hematology is not just a fancy gadget; it is a a pivotal technological transformation  that promises a more efficient and patient-focused future.  The collaboration between AI and hematologists and hematology laboratories will continue to evolve, offering significant advancements that enhance the overall quality of patient care.

References

• Radakovich N, Nagy M, Nazha A. Artificial Intelligence in Hematology: Current Challenges and Opportunities. Curr Hematol Malig Rep. 2020;15(3):203-210. doi: 10.1007/s11899-020-00575-4. PMID: 32239350.
• Walter W, Haferlach C, Nadarajah N, Schmidts I, Kühn C, Kern W, Haferlach T. How artificial intelligence might disrupt diagnostics in hematology in the near future. Oncogene. 2021;40(25):4271-4280. doi: 10.1038/s41388-021-01861-y
• Gedefaw L, Liu CF, Ip RKL, Tse HF, Yeung MHY, Yip SP, Huang CL. Artificial Intelligence-Assisted Diagnostic Cytology and Genomic Testing for Hematologic Disorders. Cells. 2023;12(13):1755. doi: 10.3390/cells12131755
• Walter W, Pohlkamp C, Meggendorfer M, Nadarajah N, Kern W, Haferlach C, Haferlach T. Artificial intelligence in hematological diagnostics: Game changer or gadget? Blood Rev. 2023;58:101019. doi: 10.1016/j.blre.2022.101019