The report is the product

A label is not an answer. The system builds its case the way a careful morphologist would — count, measure, reason, show — and then writes it down, with its limits stated on the page. The full case is laid out below, section by section, with a note on why each step matters.

About this example. This is an illustrative walkthrough of the report format, using measurements from a research case analysed with the Hemalogic.ai system. It is presented to show the structure of the output and reasoning of the model, not as evidence of diagnostic performance. The gallery image is representative morphology supplied by an expert hematopathology collaborator — it is not an image of the case specimen. Every report the system produces carries the disclaimer: a morphologic impression for pathologist review, not a diagnosis.

Peripheral blood — morphologic analysis

Research case · digital cell gallery + whole-field image · generated by Hemalogic analysis pipeline

Specimen & method

Input: Digital cell gallery and whole-field image from a digital morphology analyser
Scale calibration: Self-derived from the erythrocyte population (median diameter ≡ 7.5 µm)
Analysis: Cell-type classification with conformal certification + morphometric census
Clinical data: None supplied — morphology only

Input. The system takes the images the analyser already produced. Scale is derived from the specimen itself — the red-cell population is the ruler — so no instrument metadata is required.

Quantitative census

Measure	Result	Basis
Nucleated cells	Direct enumeration	Segmented and counted, smudge cells excluded from the differential
Leucocyte : erythrocyte ratio	Markedly increased	vs ~1:700 in a normal film — a scale-free indicator
Dominant population	Small lymphoid cells	Single population, >85% of nucleated cells

Count first, interpret second. Every figure in the report traces to a direct enumeration. Smudge cells are identified by shape and excluded from the differential so they cannot inflate the lymphoid fraction.

Morphometric findings

Feature	Measured	Interpretation
Nuclear diameter	Small, low dispersion	Monomorphous — one dominant population rather than a reactive spectrum
Chromatin clumps / nucleus	12 (median)	Coarsely clumped, condensed — a mature chromatin pattern
Internal chromatin SD	20.9	High contrast; blasts characteristically fall below 10
Nuclear-to-cytoplasmic ratio	High	Scant cytoplasm
Smudge (Gumprecht) cells	Present	Detected by shape (circularity), not by paleness

This is the discriminating step. Chromatin texture is measured, not eyeballed: the number of discrete condensed clumps per nucleus, and the internal contrast. A mature clumped pattern and a blast's fine chromatin separate cleanly on these numbers.

Morphologic impression

MONOMORPHOUS POPULATION OF SMALL, MATURE-APPEARING LYMPHOID CELLS WITH SMUDGE CELLS.
The appearances are those described in chronic lymphocytic leukaemia / small lymphocytic lymphoma (CLL/SLL).

Morphologic impression for pathologist review — not a diagnosis.

The impression names a pattern, not a patient. It is written for a pathologist to review, and it says so in the report itself.

Differential — and what the measurements exclude

Considered	Status	On what basis
Reactive lymphocytosis	Against	Population is monomorphous; reactive expansions are pleomorphic with abundant cytoplasm
Acute lymphoblastic leukaemia	Excluded on chromatin	Blasts show finely dispersed chromatin with low internal SD; measured pattern is coarse and condensed
Mantle cell lymphoma	Not excluded	Cannot be separated on morphology alone — stated explicitly rather than omitted

Recommended next steps are listed in the full report — for this pattern, flow-cytometric immunophenotyping is the discriminating test.

What it rules out — and what it cannot. The report states the entity morphology cannot exclude. Naming the limit is what makes the rest of it usable.

Cells of concern — annotated for review

Peripheral blood field: monotonous small mature lymphocytes with clumped chromatin and a smudge remnant — One field, four findings: **small mature lymphoid cells** with **coarsely clumped ("cracked") chromatin** and scant pale cytoplasm; **population monotony** — near-identical forms across the field; and a **smudge-cell remnant** (left) reflecting the fragility typical of this population. Red-cell background retained for scale.

Representative field supplied by an expert hematopathology collaborator to illustrate the features named above; it is not an image of the case specimen.

Evidence travels with the conclusion. Reviewers see the cells that drove the assessment, each with the feature it demonstrates and the reason it was selected.

Limitations — stated in every report

No clinical data or CBC accompanied this specimen; counts are areal estimates from the imaged field and carry sampling uncertainty. Morphology cannot separate CLL/SLL from mantle cell lymphoma. Where an automated count could not be supported by the image, the report says so rather than quoting a number — on this case the platelet estimate was invalidated on audit and reported as a floor.

Limitations are part of the output. Where a measurement could not be supported, the report reports that instead of a number.