Skip to content

Retrospective evaluation of a commercial AI algorithm in conventional chest and breast imaging

Retrospective evaluation of a commercial AI algorithm in conventional chest and breast imaging

Status
Active, not recruiting
Phases
Unknown
Study type
Observational
Source
DRKS
Registry ID
DRKS00031189
Enrollment
3000
Registered
2023-02-09
Start date
2022-05-26
Completion date
Unknown
Last updated
2025-04-07

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

Conditions

Diseases that indicate chest or breast X-ray imaging

Interventions

Group 1: Patients undergoing X-ray imaging of the chest or breast. X-ray images of the thorax and breast are subjected to automated, software-based diagnostic support and purely human diagnostics in v

Sponsors

Unversitätsklinikum Freiburg, Klinik für Diagnostische und Interventionelle Radiologie
Lead Sponsor

Eligibility

Sex/Gender
All

Inclusion criteria

Inclusion criteria: Retrospective inclusion of patients who have received conventional X-ray imaging of the chest or breast since January 1st, 2015.

Exclusion criteria

Exclusion criteria: A further specific patient exclusion is not planned in order to prevent falsification of the analysis for patient subgroups.

Design outcomes

Primary

MeasureTime frame
Diagnostic accuracy of the LUNIT software compared to human reading. After annotation of all image data, this is evaluated using the following statistical metrics: receiver operating characteristic curves; sensitivity, and specificity.

Secondary

MeasureTime frame
After completion of the software-supported reporting, the rate of false-positive and false-negative reports.

Countries

Germany

Contacts

Public ContactJakob Weiß

Unversitätsklinikum Freiburg, Klinik für Diagnostische und Interventionelle Radiologie

jakob.benedikt.weiss@uniklinik-freiburg.de++49 761 270-38061

Outcome results

None listed

Source: DRKS (via WHO ICTRP) · Data processed: Feb 4, 2026