Doctor & PhD-student
Sabine Morris Delhez
Department of Radiology
Project management | ||
Project status | Open | |
Data collection dates | ||
Start | 18.08.2025 | |
End | 28.02.2027 | |
This study explores whether artificial intelligence (AI) can help reduce time spent in the emergency department (ED) without compromising diagnostic quality. Missed fractures are a common error, and AI may support faster, accurate decision-making. We compare standard care with an AI-assisted workflow in two patient cohorts to assess effects on diagnostic accuracy and ED efficiency.
Emergency departments are under constant pressure to maintain patient flow while ensuring high-quality care. Missed fractures are among the most common diagnostic errors, often resulting in delayed treatment. Artificial intelligence (AI) has shown potential to support radiographic interpretation and improve efficiency. This study aims to evaluate whether an AI-assisted workflow can reduce patient time in the ED while maintaining diagnostic accuracy in fracture detection.
The study includes patients presenting to the Emergency Department at Odense University Hospital with suspected fractures. Only patients eligible for discharge based on a normal X-ray and predefined clinical criteria are included. Participants are outpatients, and both cohorts will consist of individuals assessed during routine ED visits.
No biological material will be collected. The study is based on routinely collected clinical data, including timestamps (registration, X-ray, discharge), demographic information, and radiology results. Data will be extracted from the patient journal. No data from national registries will be used.
Department of Orthopedic Surgery & Traumatology at OUH
The Emergency Department at OUH
Department of Nuclear Medicine at OUH
The Danish Centre of Clinical Artificial Intelligence (CAI-X) at OUH/SDU
Centre for Innovative Medical Technology (CIMT) at OUH/SDU
Department of Clinical IT at OUH
The Improvement team at OUH
The Center of research together with patients and relatives (ForSa-P) at OUH
Regional IT, Region of Southern Denmark