AI for fracture detection in children with osteogenesis imperfecta – Study Results | July 2025

“External validation of an artificial intelligence tool for fracture detection in children with osteogenesis imperfecta: a multireader study”

Cato Pauling, Harsimran Laidlow-Singh, Emily Evans, David Garbera, Rosalind Williamson, Ranil Fernando, Kate Thomas, Helena Martin, Owen J. Arthurs & Susan C. Shelmerdine
Publication : july 2025 | European Radiology

 

[New paper: AI + Radiologist = stronger for rare paediatric fractures] We’re proud to share the results of an study just published in European Radiology, evaluating Milvue’s AI solution in a challenging setting: fracture detection in children with osteogenesis imperfecta.

Over a 13-month period, 336 radiographic projections from 48 pediatric patients were reviewed. Radiologists achieved 83.4% accuracy on their own. With AI assistance, accuracy jumped to 90.7% compared to 74.8% for AI alone. The AI notably helped reduce false positives, while radiologists retained full decision-making control.

This work highlights the power of human/AI collaboration, especially in rare and underrepresented conditions. It reinforces our core belief: AI should support, not replace, the clinician.

> Link to the scientific journal

 

> Back to the set of Scientific Publications