In the field of hematology, the incorporation of Artificial Intelligence (AI), Machine Learning (ML), and Deep Learning (DL) models is driving revolutionary advancements. These technologies have been harnessed to develop algorithms that automate the interpretation of abnormal peripheral blood smears, classify blood disorders, predict disease prognosis, and identify new biomarkers.

AI/ML applications in laboratory settings have been explored extensively. These technologies have been used to create algorithms that predict laboratory test values, enhance laboratory efficiency, automate laboratory procedures, support precise interpretation of laboratory tests, and refine laboratory medicine information systems—some achieving notable accuracy.

Overall, AI/ML technology shows potential in utilizing extensive medical data to generate more individualized and personalized interpretations of test findings. Although there has been some adoption of AI and ML methods in the laboratory environment, particularly in molecular pathology (for instance, categorizing central nervous system tumors using DNA methylation profiling) and digital pathology (such as image analysis), progress has been incremental.

Traditional approaches, such as manual examination of peripheral blood smears (PBSs), continue to serve as important diagnostic tools. While they offer valuable insights into various disorders such as leukemia, anemia, infections, and allergies, these  routine microscopic examinations in the hematology laboratory are time-consuming, error-prone, and inefficient, particularly in healthcare settings with high patient volumes.

Fuelled by the rising incidence of hematologic disorders and the complexities and challenges presented by global pandemics, AI in hematology offers significant and substantial possibilities.ML algorithms have been used to predict the likelihood of thrombosis in individuals with myeloproliferative neoplasms (MPNs) and to discover novel subtypes of leukemia using genomic information.These algorithms have also been applied to analyze extensive datasets of hematology images, like PBSs, bone marrow aspirates (BMAs), and lymph node biopsies. This has led to the development of AI-supported diagnostic systems, helping specialists in hematology and pathology  achieve more precise diagnoses.

Given the dynamic and changing nature of hematologic disorders and the unique characteristics of different cells, introducing AI and ML methodologies brings about a rapid, more precise, and efficient analysis of the peripheral blood smear. These technologies aim  to reduce turnaround time for patient diagnosis, reduce interpersonal variability, improve healthcare delivery, lower healthcare costs, and predict potential patient prognosis.

Traditional “cell-locating” digital systems cannot show individual cells within the clinical context of the sample, presenting challenges for laboratory scientists and physicians who need a full slide overview before zooming in on specific cells or areas of interest. These systems are mainly used for initial screening, requiring further manual examination, particularly for identifying abnormalities like leukemia or suspected pathological cell types.

In contrast, Full-Field digital cell morphology represents a novel digital imaging approach that utlizesintegrates computational photography to capture all relevant regions of clinical interest both the overall sample and intracellular details simultaneously. Departing from the reliance on precise mechanics and costly optics, this approach employs a unique physics-based model to produce sharp, high-quality digital images. Unlike conventional systems that generate snapshots, Full-Field cell morphology allows laboratory experts to zoom in on any cell or group of cells at 100X magnification, while retaining the larger sample context. This comprehensive viewing capability ensures that crucial details can be examined anywhere on the scan, including the monolayer and feathered edge. Moreover, this technology facilitates seamless operation across hospital and lab networks of varying sizes, enabling workload balancing, remote consultations, and addressing personnel shortages effectively.

The synergy between AI, ML, and DL models and traditional hematology diagnostics holds enormous promise and potential  for the future. As the progress unfolds, overcoming challenges and embracing innovative technologies will undoubtedly usher in  a new era in hematology, where accuracy, efficiency, and patient care take center stage.

References

• 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
• Blum K. A Status Report on AI in Laboratory Medicine. Association for Diagnostics and Laboratory Medicine. 1 January 2023. Available from https://www.myadlm.org/cln/articles/2023/janfeb/a-status-report-on-ai-in-laboratory-medicine?utm_source=cln-email&utm_medium=email&utm_c%E2%80%A6 [Accessed 16 January 2024]
• Katz BZ, Feldman MD, Tessema M, Benisty D, Toles GS, Andre A, et al. Evaluation of Scopio Labs X100 Full Field PBS: The first high-resolution full field viewing of peripheral blood specimens combined with artificial intelligence-based morphological analysis. Int J Lab Hematol. 2021 Dec;43(6):1408-1416. doi: 10.1111/ijlh.13681