Organize the worklist
Filter cases by laboratory, find outstanding reviews and see which image sets need attention.

BLOOD MORPHOLOGY
HAEMARA is developing an AI-assisted blood-morphology platform for hospitals and laboratories. The research beta brings images, blood-count context, reviewer decisions and test reports into one workflow—so laboratory teams can explore how a case moves from intake to approval.
Beta with test cases and public research images. Not for clinical use.
Images → Review → Report
THE WORKFLOW
Filter cases by laboratory, find outstanding reviews and see which image sets need attention.
Inspect reference cell images alongside blood counts. Record morphology, notes and second-review decisions.
Produce a test report with reviewed findings, the original blood count and a record of approval. Export a table or print a report.
WHY HAEMARA
Bring image context, blood counts, reviewer decisions and research exports into one focused workflow. Explore what the beta already supports—and the flexible, source-aware review workspace we aim to prove with research teams.
Explore the benefits and our approach →ONE CASE, CONNECTED DECISIONS
Move from the worklist to white-cell review, red-cell morphology and platelet checks, with test blood-count values available in context. Save a draft and return to outstanding decisions.
Explore a second-review queue with the first reviewer’s note. Test how unresolved classifications and quality holds affect approval.
Explore institution filters, case status and completed test reports. Inspect how decisions are retained in a report snapshot and exported or printed.
The beta uses selectable test reviewers. Role-based authentication and production access controls remain to be implemented.
RESEARCH SOFTWARE
Explore public microscopy images, record researcher-entered cell labels and preserve notes and review decisions. This work supports reference-image research and workflow evaluation.
Open image reviewUse prepared cases to explore worklists, blood-count context, quality holds and report layouts. The demonstration keeps generated case values separate from the source public-image records.
Try a test caseThe planned release module will organize versioned image sets and annotations for controlled research handoffs. Its development builds on the beta’s review records and dataset provenance work.
INSTITUTIONAL EVALUATION
Potential projects include image-review feasibility, reference-label preparation and evaluation of research workflows. Each project will specify its image quantities, deliverables, permitted data uses and participating reviewers.
Institutions retain physical samples and slides. Any research-image access will follow an agreed scope, data permissions and applicable institutional review. Initial enquiries should contain professional contact information and workflow interests only.
Discuss a research scopeEXPLORE THE BETA
The workflow includes 220 prepared cases. Their blood counts and classifications are constructed test data; public white-cell photographs are illustrative and not patient-matched. Red-cell and platelet views are schematic.
Analysis and laboratory-system delivery are simulations. The beta demonstrates review and reporting behavior; it does not perform trained AI inference or establish clinical performance.
IN DEVELOPMENT
Following the initial development of haemara-ai.com, HAEMARA built haemara.app with updated branding and an interactive beta experience.
The beta demonstrates the workflow around a future morphology engine. It includes 220 prepared test cases, a public-image workbench and a discovery form for institutional discussions.
The intended service uses authorized digital images. Blood samples and physical slides stay with the institution. Scanner compatibility, model performance and clinical use require separate evaluation and approvals.
Explore the public-image workbench →ADDITIONAL RESEARCH TOOLS
Alongside the laboratory workflow beta, the public-image workbench lets you create a named research batch, record its purpose and select images from the bundled PBC and BCCD collections. Original source labels stay separate from human assessments.
Group images for a specific research question while retaining each source’s identity and annotation format.
Inspect source images, record assessments and see which selected images still need review.
Download a dated JSON manifest or CSV with source identifiers, image hashes, licence metadata and current assessments. Image pixels are not included.
An additional research workflow, saved in this browser. Institutional image imports, authenticated shared review and adjudication remain development work.
Research batches NEW →Explore the beta, then tell us about your review process and evaluation priorities.