Automated hematology is at the center of modern diagnostics, providing fast and reliable complete blood counts (CBCs) that support decisions across internal medicine, emergency care, oncology, and more. Over the last decade, the field has shifted from simple impedance‑based counters to AI‑powered analyzers that deliver detailed morphology and decision support along with numeric results.

This article explains how automated hematology works in 2026, what has changed with image‑based AI, and how to choose the right hematology auto analyzer for your lab or clinic.

What automated hematology actually does

A hematology auto analyzer automates the measurement of key blood parameters, including:

• White blood cell (WBC) count and differentials
• Red blood cell (RBC) count and indices (MCV, MCH, MCHC, RDW, etc.)
• Hemoglobin (HGB) and hematocrit (HCT)
• Platelet (PLT) count and platelet indices

Beyond these basics, modern automated hematology analyzers often provide additional features:

• 3‑part, 5‑part, or 7‑part WBC differentials to separate neutrophils, lymphocytes, monocytes, eosinophils, basophils, and sometimes immature or abnormal forms.
• Advanced ratios and indices, such as neutrophil‑to‑lymphocyte ratio (NLR) and platelet‑to‑lymphocyte ratio (PLR), which are increasingly used in prognosis and risk stratification.
• Morphology flags and digital images, especially in AI‑enabled systems, which help detect abnormal cells and guide further review.

Ozelle’s EHBT series, for example, combines CBC parameters with digital imaging and AI morphology (Complete Blood Morphology, CBM) in a single automated run.

From impedance to AI: the evolution of automated hematology

Historically, hematology analyzers have evolved through four major stages:

Stage	Core technology	Strengths	Limitazioni
Manual microscopy	Visual smear examination	Rich morphology, expert insight	Slow, subjective, labor‑intensive
Impedenza elettrica	Cell volume & count	Simple, affordable, robust	Limited to size and count; little morphology
Flow cytometry	Light scatter & fluorescence	Detailed classification	Complex, expensive, marker‑dependent
Image‑based AI	Digital imaging + AI	Rich, standardized morphology at scale	Requires strong optics and compute

Early analyzers automated only counting and sizing; today’s AI‑powered systems combine digital imaging, multispectral optics, and deep learning to approach expert‑level morphology in an automated, consistent way.

How an AI‑powered hematology auto analyzer works

Using Ozelle’s EHBT‑50 as an example, a typical automated hematology workflow goes far beyond simple counting.

Step 1: Sample loading

A small amount of venous or capillary blood—often 30–100 µL—is loaded into the analyzer or into a dedicated cartridge.

Step 2: Automated pretreatment

The analyzer automatically handles dilution, mixing, staining, and preparation, which reduces manual steps and standardizes processing.

Step 3: Imaging and signal acquisition

High‑resolution cameras capture multiple images of cells, often at several focal planes and wavelengths, providing a detailed view of cell shape, granularity, and nuclear morphology.

Step 4: AI morphology analysis

Convolutional neural networks (CNNs) and other AI models segment and classify cells into various WBC, RBC, and PLT categories, including abnormal forms. These models are trained on large datasets of real clinical images.

Step 5: Parameter calculation

The system computes standard CBC parameters plus extended morphology‑based indicators and flags, which can total 30–40+ parameters in advanced analyzers.

Step 6: Reporting and integration

Results are displayed on the analyzer screen, can be printed, and are typically transmitted to LIS/HIS or a cloud platform for storage and review. Some systems also generate suggested interpretive comments based on parameter patterns.

For many routine samples, this entire process takes around six minutes per test.

Clinical benefits of automated hematology with AI morphology

AI‑enabled automated hematology delivers several clear clinical and operational benefits:

Better abnormal cell detection

AI models trained on millions of cells can detect subtle morphological changes and abnormal populations that can be difficult to identify consistently with manual review alone. This includes immature granulocytes, atypical lymphocytes, abnormal platelets, and shape changes in red blood cells.

More consistent and reproducible results

Manual smear interpretation can vary between technicians and even for the same person on different days. Automated imaging and AI classification apply the same decision rules every time, helping to reduce inter‑observer variability and improve consistency.

Faster turnaround time

By delivering both CBC and morphology in a single automated run, AI analyzers significantly reduce the need to prepare and review separate smears. This shortens turnaround time, which is particularly important for emergency and outpatient settings.

Smarter triage for smear review

Automated systems can flag only those samples that most truly require manual review, allowing hematology experts to focus on high‑value cases. This improves overall efficiency while maintaining quality standards.

How automated hematology ties into workflows and connectivity

Modern hematology auto analyzers are increasingly integrated into wider digital ecosystems:

• LIS/HIS integration Direct data exchange with laboratory and hospital information systems reduces transcription errors and speeds reporting.
• IoT and remote monitoring Cloud platforms allow remote monitoring of analyzer status, consumables, and performance metrics across multiple sites.
• Analytics and quality management Aggregated data from hematology analyzers can be used for quality control, capacity planning, and even clinical research.

Ozelle, for example, provides a cloud platform that connects EHBT analyzers and other devices, allowing distributors and healthcare networks to monitor fleets, manage consumables, and analyze testing trends.

Choosing the right hematology auto analyzer

When selecting an automated hematology system, consider the following questions:

1. What level of WBC differential do we need?
   1. 3‑part analyzers are often sufficient for primary care.
   2. 5‑part and 7‑part analyzers are preferred for hospitals and specialized centers that manage complex hematology and oncology cases.
2. Do we need AI morphology and digital images?
   1. AI morphology reduces smear workload and provides more detailed information.
   2. Access to images can be valuable for training, case review, and consultation.
3. What is our current and projected test volume?
   1. High‑volume labs gain greater benefit from automation and AI because time savings and reduced smear rates accumulate quickly.
4. Do we need additional test channels in one device?
   1. Multi‑functional analyzers that combine hematology with immunoassay or basic chemistry can simplify the lab setup and support broader diagnostic workflows.
5. What about connectivity and long‑term support?
   1. Look for systems designed to connect to LIS/HIS and cloud platforms, and for vendors that provide remote service and software updates.

You can see detailed examples of different automated hematology solutions on Ozelle’s website at https://ozellemed.com/en/.

FAQs – Automated Hematology

Q1. Is automated hematology accurate enough to replace all manual smears?

AI‑powered analyzers greatly reduce the number of smears required, but manual review remains important in particularly unusual, complex, or critical cases.

Q2. What is the difference between a basic hematology analyzer and an AI‑powered one?

Basic analyzers mainly count and size cells, while AI‑powered systems also analyze morphology through imaging, which improves abnormal cell detection and reduces manual workload.

Q3. Do smaller clinics really need AI morphology?

Many smaller clinics benefit from AI if they want faster, more detailed results without building a large lab team; however, very low‑volume sites may prioritize upfront cost and basic CBC only.

Q4. How do automated hematology analyzers fit into multi‑site networks?

Connected analyzers allow centralized monitoring of performance, easier quality management, and more consistent results across locations.