Research Ethics Compliance
How this research programme treats patient data, research ethics and the governance of its rules.
Last updated:
This page is published in several languages. Where the versions differ, the English text is the authoritative one.
1. Scope
BlockEnergy ACR is a research programme run by Cornerstone Research International Ltd ("CRIL", registered in Ireland, number 386943) with University College Dublin, Ireland, and Zhengzhou University, People's Republic of China. This statement covers the programme's research conduct. How personal data is handled is set out separately in our Privacy Policy.
2. Principles we work to
- Patient data stays where it is. The programme is built so that clinical records never leave the institution that holds them. Only rules and model updates travel between sites — the principle we describe as “Data Stays. Rules Travel.”
- Minimum data. The reasoning uses a small set of clinical values, and nothing that identifies a patient. Direct identifiers are not accepted by the mobile application at all.
- The clinician decides. The system produces reasoned suggestions with the rules that fired and the evidence behind them. It never decides, and a qualified clinician is responsible for every clinical decision.
- Reasoning you can inspect. Results come from a published, deterministic rule set rather than an opaque model, and the reasoning path is shown with every result. Where a statistical layer adds a confidence figure, it is advisory and labelled as such.
- Local law first. Each participating site works to its own national rules — the GDPR in Ireland and the European Union, the PIPL and related rules in the People’s Republic of China — rather than to a single exported standard.
3. No patient data on this website
Everything demonstrated on this website runs on invented cases. The demonstration database holds synthetic records that describe no real person, and the reasoning shown in public pages never touches a clinical system.
Visitors must never enter real patient data, or any information that could identify a real person, on any page of this website.
4. Work with real clinical data
Any work involving real clinical records is carried out by the clinical partner, on that institution's own infrastructure, under that institution's ethics approval and national regulatory arrangements, and under a written agreement with CRIL. CRIL does not receive identifiable patient data.
The present stage of the programme is a technical evaluation by invited reviewers using synthetic cases. It is not a clinical trial, and no clinical study using this software has begun.
5. Governance of the rules
Clinical rules are not changed informally. A change is proposed, reviewed and recorded, and the record of who proposed it, who approved it and when is kept as an audit trail, so that any result can be traced back to the exact rule set that produced it.
- A steering committee, co-chaired by the clinical leads at the partner institutions, oversees the programme.
- Specialist workgroups cover imaging AI, molecular diagnostics and implementation science.
- Each release records the ontology version and the rule counts it ran with, and the software checks that what ran matches what was published.
6. Status and limitations
This software is a research prototype. It is not a medical device, it carries no regulatory clearance or CE marking, and it must not be used to care for a patient.
The rule set is limited to breast cancer decision support and is incomplete by design. Its suggestions have not been validated against clinical outcomes. Published results describe how the reasoning behaves, not how patients fare.
7. Raising a concern
If you have a concern about the conduct of this research, or about anything published on this website, write to us and we will reply within one month: info@acragent.com
A concern about work carried out at a participating hospital or university may also be raised directly with that institution's own research ethics office.