Haem.io

Tools

Clinical workflow tools

A small directory of standalone haem.io tools. These are intended for structured exploration, audit, and decision support, not autonomous clinical decision-making.

Data handling note

Public hosted demos should not be used with identifiable patient data. Hospital or trial workflows should use an approved local deployment where required by information governance.

ToolUseStatusData handlingAccess

ELN 2022 AML Risk Calculator

Risk stratification calculator

Apply structured ELN 2022 intensive-treatment AML risk logic from genetic and cytogenetic inputs, with visible reasoning.

Available

Uses the haem.io classifier proxy with local deterministic fallback if the backend is unavailable. Do not enter patient identifiers.

Open

MDS / AML Boundary Checker

WHO/ICC classification helper

Compare WHO 2022 and ICC 2022 outputs around the MDS, MDS/AML, and AML boundary using structured inputs.

Available

Uses the haem.io classifier proxy with local deterministic fallback if the backend is unavailable. Do not enter patient identifiers.

Open

TP53 Multi-Hit Checker

Molecular rule checker

Check how TP53 configurations affect WHO 2022, ICC 2022, and ELN 2022 outputs.

Available

Uses the haem.io classifier proxy with local deterministic fallback if the backend is unavailable. Do not enter patient identifiers.

Open

AML Treatment Explorer

Treatment pathway explorer

Step through the Coats-Delphi AML treatment pathway cases using structured molecular and clinical inputs.

Available

Browser-based use. No account required for the public explorer.

Open

ELN AML Response Calculator

Response criteria calculator

Apply ELN 2017 and ELN 2022 AML response criteria from marrow findings and blood counts, with visible reasoning and PDF report output.

Demo hosted

Hosted demo. For patient-identifiable use, deploy the standalone Node app inside the relevant organisation.

Open

Local deployment

The ELN AML Response Calculator is available as a lightweight Node app for private server deployment. It has no LLM component and is designed to show the response criteria and reasoning used for each assessment.

Request deployment pack