Bone Marrow Aspirate

What it does

Bone Marrow Aspirate reads a May-Grünwald-Giemsa-stained aspirate smear and produces the myelogram, the marrow differential that otherwise takes a trained morphologist several hundred manual cell counts. The blast percentage in that count is one of the values that decides how a myeloid neoplasm is classified.

The run finds the well-spread areas of the smear, classifies each nucleated marrow cell into one of about twenty classes across the myeloid, erythroid, lymphoid, monocytic and plasma-cell lineages and blasts, and reports the differential in percent, the myeloid-to-erythroid (M:E) ratio and the blast percentage. Cells with dysplastic features are flagged, and every class has its own cell gallery, so each number can be checked against the cells behind it.

Intended use

Decision support for the haematologist or haematopathologist reporting an aspirate: a full cell-by-cell count to review, and a blast percentage set against the WHO-HAEM5 and ICC 2022 thresholds.

Who it is for

Haematologists, haematopathologists, haematology trainees and laboratory scientists in marrow-reporting labs. Developers use the run to add structured myelograms to a laboratory system or to a leukaemia registry.

Accepted input

• Whole-slide image of an aspirate smear, as DICOM VL Whole Slide Microscopy Image or a scanner file (SVS, NDPI, MRXS, CZI, OME-TIFF and others).
• Cell images: JPEG, PNG or TIFF crops of single marrow cells from a digital morphology system.
• One aspirate per run.

Requirements

Requirement	Why
May-Grünwald-Giemsa stain	Cytoplasmic granules and chromatin define the marrow classes
Thin, well-spread areas on the smear	Cells are counted where they lie singly, not in thick particle trails
Pixel spacing on whole slides	Cell size separates blasts, promyelocytes and lymphocytes

Optional context

The blood count, the peripheral smear result, prior diagnosis or treatment and the clinical question can be passed in clinical_context; the guidance uses them.

The result

Section	Content
Marrow differential	Class of every counted cell, differential percentages, M:E ratio, blast percentage, dysplasia flags and a cell gallery
Draft report	Opt-in (options.report): an English narrative that interprets the result for the reader (findings, impression, limitations); every number is checked against the findings

Standards

• Classification: blast thresholds of the WHO Classification of Haematolymphoid Tumours, 5th edition (WHO-HAEM5), and the International Consensus Classification (ICC 2022).
• DICOM: classified cells as Microscopy Bulk Simple Annotations on the slide, the myelogram in SR TID 1500, an Encapsulated PDF.
• Interoperability: FHIR R4 DiagnosticReport with LOINC-coded Observation entries for the differential.

Counting notes

A marrow differential is conventionally based on at least 300 to 500 nucleated cells, which is why it is slow by hand. The run classifies every countable cell in the well-spread areas, and the guidance places the blast percentage against the thresholds of both WHO-HAEM5 and ICC 2022 instead of choosing one framework.

Works with

A blast alert from Peripheral Blood Smear (path/blood-smear) is the usual reason to run this model.