PROJECT TEAM
Lone Kjeld Petersen, Professor, Chief Physician, MD, OUH, Department of Gynaecology and Obstetrics
Esmaeil Nadimi, Professor, Head of Unit, SDU, Applied AI and Data Science
Peter Børker Nielsen, Head of Unit, OUH, Centre for Clinical Artificial Intelligence
THE NEED
Colposcopy is a subjective and skill-dependent procedure. In Denmark, guidelines require four biopsies to ensure diagnostic quality, which causes discomfort for patients and a heavy workload for clinicians and pathologists. Less experienced colposcopists may struggle with accurate targeting, making the procedure even more challenging.
THE SOLUTION
PEACE uses artificial intelligence trained on a large, high-quality dataset of annotated colposcopy videos. It provides real-time visual guidance during the procedure by overlaying highlighted biopsy targets on the live video feed. This helps clinicians identify the most relevant areas for biopsy, reducing the number of samples needed and improving the consistency of care.
PEACE - Prediction of cell changes in the cervix using artificial intelligence - Odense University Hospital
Call 3 - 2022
500.000 DKK
Clinical Area
Gynecology , Cancer Screening
Technology
AI-guided diagnostic support
PROJECT SUMMARY
PEACE is an AI tool that guides biopsy placement during colposcopy in real time. The solution integrates directly with existing colposcopy equipment and highlights areas of concern on screen, helping clinicians reduce the number of biopsies, without compromising diagnostic accuracy.
CLINICAL IMPACT
By reducing biopsies from four to one, PEACE minimizes physical discomfort and emotional stress for the 15,000 Danish women referred for colposcopy each year. It also shortens analysis time for pathologists, eases workload for healthcare staff, and improves access to high-quality care—even in clinics with limited experience. This leads to faster diagnosis, earlier treatment, and a more uniform standard of care across the system.