Research Organisations

For Research Organisations

Pharma, clinical, and computational-biology teams see a gap between the richness of real-time monitoring data and the ability to obtain it in a reproducible, auditable form. ICU episode data is collected for clinical purposes, not for research: it lacks the provenance guarantees a trial protocol requires, and system state at the moment of a decision is rarely preserved.

The problem we solve

Research teams need:

1 · A reproducible runtime

Replaying the same clinical episode must produce the same result.

2 · A data-provenance chain

Preserved under non-repudiable audit.

3 · Visualisations

Correlating signals, decisions, and outcomes over time.

Commercial systems do not provide any of these as standard.

Our answer

HealthOS ICU provides three capabilities relevant to research:

Deterministic replay

The Compute core (the bedside execution module) is replay-safe: load any historical waveform segment with the version of the compiled bundle (rule version) active at that time, and you get the exact same output the live system produced. This is the foundation for both retrospective analysis and prospective controlled experiments.

Evidence-of-fact receipt

Every output is backed by an evidence-of-fact receipt (proof of provenance): a cryptographic chain from the approved knowledge graph through SHACL validation, the knowledge compiler, and runtime execution to the specific recommendation. This receipt is publishable as a traceable citation.

Retrospective episode review

The diagnostic contour (the closed episode-review cycle) is logged in the audit event log and visualisable via DICOMweb segment retrieval. Research teams can replay an episode, annotate decisions, and export structured data for statistical analysis.

Diagnostic contour Five links in a circle: evidence, classify, decide, remediate, close; an arrow returns from close to evidence; every step is recorded in the audit event log. Evidence Classify Decide Remediate Close every step is recorded in the audit event log

Collaboration model

Joint study

The research organisation provides the study protocol; the platform provides a deterministic runtime, evidence-of-fact receipts, and structured data export.

Data partnership

Retrospective episode cohorts are provided through a governed export from the audit log; de-identification is part of the governance framework.

Methodological co-authorship

The closed knowledge-distillation cycle methodology is citable; co-authorship on methodology papers is possible.

Documents and training

Research delivery options, including P-RES-AGG and P-RES-OPS, are entitlement-bound add-ons. The Program Certificate remains evidence of training; fulfillment goes through a separate admission review.


Discuss a joint study or a data partnership.

Contact us about research