What are the factors limiting automation in haematology?What are the factors limiting automation in haematology?

Laboratory haematology has undergone remarkable transformation over the past decades, shifting from predominantly manual processes to highly automated systems capable of analysing thousands of samples daily with precision and reproducibility (Briggs et al., 2019). The evolution of automated haematology analysers has fundamentally changed laboratory workflows, diagnostic capabilities, and ultimately patient care. This article examines the applications of automation in modern haematology laboratories while acknowledging the limitations that continue to necessitate human expertise.

Applications of automation in haematology

Modern haematology analysers employ multiple technologies including flow cytometry, electrical impedance, and optical scatter to generate comprehensive blood counts with exceptional precision (Simson et al., 2021). These instruments can analyse thousands of cells per sample within seconds, providing quantitative measurements of red blood cells, white blood cells, platelets, and associated parameters including cell size, hemoglobin concentration, and cellular distribution metrics.

Beyond standard parameters, advanced analysers can identify immature or abnormal cell populations through specialised flags and scatter plots, prompting additional testing or microscopic review when necessary (Green & Wachsmann-Hogiu, 2022). This capability serves as an effective screening mechanism for detecting various haematological disorders, from common anaemias to malignant conditions.

Related:AI is rewriting the rules of in vitro fertilisation

Laboratory automation extends beyond individual analysers to encompass entire workflow systems. Track-based automation connects pre-analytical processing (specimen reception, centrifugation, aliquoting) with analytical platforms and post-analytical handling (storage, disposal), minimising manual intervention and reducing turnaround times (Lippi & Da Rin, 2019). These integrated systems incorporate barcode tracking, automated specimen routing, and electronic result verification, significantly decreasing the potential for human error while enhancing traceability.

Quality management has been revolutionised through automation, with continuous internal quality control monitoring and automated delta checks comparing current results with previous patient values. This facilitates immediate identification of analytical deviations or significant clinical changes requiring intervention (Plebani, 2020).

Limitations of automation

Automated systems remain limited in their ability to perform nuanced morphological assessments. The identification and characterisation of subtle cellular abnormalities — such as toxic granulation in neutrophils, reactive lymphocytes, or early leukemic blasts — often requires expert human review. While digital imaging systems have improved automated morphological analysis, they have not fully replicated the interpretative capabilities of experienced haematology professionals.

Related:Building biotech hubs: The Middle East’s biomanufacturing evolution

Automated analysers are optimised for detecting common haematological parameters and abnormalities but may miss rare conditions or unusual presentations. Hereditary erythrocyte disorders, certain parasitic infections, and some myelodysplastic syndromes can present with subtle changes that fall outside standard flagging algorithms (Constantino, 2019). These limitations necessitate supplementary testing approaches and expert microscopic review.

Automated systems rely on pre-defined algorithms that, while sophisticated, cannot adapt to all biological variations. Sample interference from lipemia, hemolysis, agglutination, or cold agglutinins can produce erroneous results that require human identification and intervention (Zandecki et al., 2021). Furthermore, certain patient populations — including neonates, oncology patients, and those with severe systemic illnesses — may present with cellular characteristics that confound standard algorithms.

Related:Laboratory plays critical role in revolutionising Alzheimer’s care

Future directions: AI and machine learning

Artificial intelligence and machine learning represent promising avenues for addressing current limitations in automated haematology. These technologies can enhance pattern recognition capabilities, enabling more sophisticated morphological assessments and potentially identifying subtle disease markers invisible to conventional analysis (Kumar et al., 2022). Through continuous learning algorithms, these systems may eventually adapt to unusual presentations and rare conditions that currently challenge automated detection.

Conclusion

Automation has undeniably transformed haematology laboratory practice, bringing unprecedented efficiency, reproducibility, and analytical capacity. However, significant limitations persist in areas requiring complex morphological interpretation and the detection of rare conditions. The integration of artificial intelligence with established automation technologies offers promising solutions, though human expertise remains essential for comprehensive result interpretation and clinical correlation. The optimal haematology laboratory of the future will likely combine advanced automation with human expertise in an integrated, collaborative diagnostic approach.

References available upon request.

Professor M A Muhibi, from the Department of Medical Laboratory Science, Faculty of Applied Health Sciences at Edo University Iyamho, will speak at the Haematology and Blood Transfusion conference at WHX Lagos Labs.