Haem provides a clinical‑grade AI platform that parses free‑text haematology reports to extract mutations, variant allele frequencies, cytogenetic abnormalities, blast counts, and flow‑cytometry markers with 100 % gene detection accuracy. The extracted data are displayed alongside the original text for transparent review, then classified using WHO 2022 and ICC 2022 criteria and automatically scored for ELN 2022/2024 risk, delivering traceable reasoning and evidence‑based decision support for hematologists, pathologists, and lab teams.
Funding
Funding not disclosed
Founders
Product
Problem
Clinicians and laboratories must manually extract and interpret complex haematology data from free‑text reports, a process that is time‑consuming, prone to transcription errors, and difficult to align consistently with evolving WHO, ICC, and ELN classification frameworks.
Solution
Haem.io offers an AI‑driven platform that parses any haematology report—morphology, molecular, cytogenetics—and automatically extracts key findings such as mutations, variant allele frequencies, cytogenetic abnormalities, blast percentages, and flow‑cytometry markers with 100 % gene detection accuracy. The extracted data are presented side‑by‑side with the original text for transparent review and correction. Using dual‑framework logic, the system classifies cases according to WHO 2022 and ICC 2022 criteria, providing a traceable reasoning chain for each decision node. It then calculates ELN 2022 (intensive) and ELN 2024 (non‑intensive) risk scores, including median survival estimates and full audit trails, to support evidence‑based treatment decisions.
Target Audience
Primary users are hematologists, clinical pathologists, and laboratory teams that need rapid, accurate classification and risk assessment of leukemia and myelodysplastic syndromes, as well as pharmaceutical researchers conducting genotype‑driven trials.
Features
- LLM‑powered extraction of mutations, VAF, cytogenetic abnormalities, blast counts, and flow markers from any free‑text haematology report
- Side‑by‑side display of parsed data versus original report for easy verification and editing
- Dual‑framework classification against WHO 2022 and ICC 2022 with fully visible reasoning chains
- Automated ELN 2022 and ELN 2024 risk stratification with median survival projections and auditable calculation steps
- Detection of TP53 allelic status (LOH, del17p, multi‑hit) and assessment of cytogenetic complexity
- Interactive sandbox for real‑time manipulation of genetic and cytogenetic inputs to observe classification outcomes
- Integrated learning modules and practice cases for continuous clinician education