Skip to content

HAVEN-AI: A Prospective Clinical Validation Study of a Privacy-Preserving Locally Deployed Medical Imaging Artificial Intelligence Appliance in Secure Hospital Edge Networks

HAVEN-AI: A Prospective Clinical Validation Study of a Privacy-Preserving Locally Deployed Medical Imaging Artificial Intelligence Appliance in Secure Hospital Edge Networks

Status
Active, not recruiting
Phases
Early Phase 1
Study type
Observational
Source
ChiCTR
Registry ID
ChiCTR2600130068
Enrollment
Unknown
Registered
2026-08-14
Start date
2026-08-17
Completion date
Unknown
Last updated
2026-08-24

For informational purposes only — not medical advice. Sourced from public registries and may not reflect the latest updates. Terms

Conditions

Common diseases in oncology and internal medicine

Interventions

CT group:None
MRI group:None

Sponsors

Dazhou Central Hospital
Lead Sponsor

Eligibility

Sex/Gender
All
Age
18 Years to No maximum

Inclusion criteria

Inclusion criteria: 1.Age >= 18 years old, gender is not limited; 2.For routine clinical diagnosis and treatment needs, undergo CT, MRI or DR examinations, and the examination sites, modalities and disease scenarios comply with the preset applicable scope; 3.The DICOM images have the basic integrity and image quality required for AI inference, doctor interpretation and expert evaluation; 4.It is allowed to use the images, final reports, necessary clinical information and timestamps generated by routine diagnosis and treatment in a safe environment approved by the hospital for de-identification research; 5.If the ethics committee approves the use of informed consent waiver or simplified procedures during a specific silent period, it shall be strictly implemented in accordance with the approved scope. 6.Be able to complete the necessary outcome verification or subsequent follow-up as required by the research.

Exclusion criteria

Exclusion criteria: 1.Severe image loss, obvious artifacts, format incompatibility, or technical problems that cannot pass quality control; 2.In the state of emergency rescue, it is impossible to ensure full informed consent and there is no qualified legal representative, or it belongs to vulnerable/special populations not covered by the protocol; 3.The subject clearly refuses, withdraws informed consent, or their data is legally not allowed to be used in this study; 4.The researcher believes that inclusion may interfere with routine medical treatment, increase unreasonable risks, or unable to obtain reliable primary endpoint evaluation.

Design outcomes

Primary

MeasureTime frame
The major clinical error rate of the final issued report;

Secondary

MeasureTime frame
The number of incidents of patient-level data leakage;specificity;The proportion of successful local forward-looking verification in the Privacy Restriction Center;missed diagnosis rate;sensitivity;Outbound data compliance rate;

Countries

China

Contacts

Public ContactZeng Fanxin

Dazhou Central Hospital

fanxinly@163.com+86 818 2379482

Outcome results

None listed

Source: ChiCTR (via WHO ICTRP) · Data processed: Sep 19, 2026