Datasets
De-identified ICU cohorts, waveform archives, ventilation logs.
Science grows when it is shared.
YODA believes that the future of intensive care depends on transparent methods, reproducible science, shared resources, and global collaboration.
Transparency is not an aesthetic. It is the mechanism by which science corrects itself. When methods are visible, errors become findable. When datasets are shared, results become testable. When code is open, conclusions become reproducible.
Intensive care is a discipline of consequence. Decisions unfold in minutes; their weight lasts a lifetime. A field that carries this responsibility cannot afford closed knowledge, hidden pipelines, or unverifiable claims.
Collaboration multiplies rigour. A protocol reviewed across continents is stronger than a protocol read within a single institution. A model validated in ten health systems is safer than a model tuned on one.
Every dataset shared has the potential to save lives.
Datasets, protocols, code, models, simulations, calculators, algorithms, and teaching material — versioned, licensed, cited.
De-identified ICU cohorts, waveform archives, ventilation logs.
Bedside protocols with full methodological transparency.
Reproducible R and Python pipelines for critical care analytics.
Interpretable models with training cards and validation reports.
Lung, circulation, and ventilator interaction simulators.
Physiology-grounded bedside calculators, open source.
Decision pathways with references, limitations, and provenance.
Slide decks, figures, and teaching modules — freely reusable.
A clinical problem, sharply framed.
Pre-registered. Peer-reviewed. Public.
FAIR: findable, accessible, interoperable, reusable.
Versioned code. Deterministic pipelines.
External cohorts. Independent replication.
Open access. Full transparency of methods.
Back to the bedside — where knowledge belongs.
A growing suite of open, physiology-grounded tools — free to use, inspect, and improve.
Every resource — findable, accessible, interoperable, reusable.
Clear stewardship, provenance, and lifecycle for every dataset.
Rigorous de-identification. Patient dignity is non-negotiable.
IRB oversight, informed consent, transparent conflicts of interest.
Creative Commons for content · permissive OSI licences for code.
Persistent identifiers — DOI, ORCID, ROR — for every artefact.
Contribute a de-identified cohort, waveform archive, or registry.
Share a bedside or research protocol for community review.
Publish code, models, or notebooks with a DOI.
Slides, figures, or modules for open teaching.
Help us fix errors, ambiguities, or reproducibility gaps.
Join the editorial and stewardship community.
Universities, hospitals, research groups, and open-source contributors — one shared infrastructure for critical care knowledge.
Counters activate as the commons grows.
"Knowledge hidden cannot improve patient care."