Purpose or Case Report: Wrist trauma is common in children and generally requires radiography for diagnosis. Many children who receive radiographs do not have fractures and are thus subjected to unnecessary radiation exposure along with increased wait times in the emergency department (ED). Ultrasound (US) is safe, cost-effective, portable and sensitive in visualizing cortical disruption, potentially making it a valuable tool for bedside assessment of fractures. This study aims to determine the feasibility of using US to detect distal radius fractures (DRF) in children and to contrast the accuracy of hand-held device compatible 2D transducers to 3D transducers that can only be used with traditional US machines. In order to address difficulties in US image interpretation by inexperienced users, we investigated the utility of an artificial intelligence (AI) network.
Methods & Materials: 127 children presenting to the ED with wrist trauma were scanned with 2D and 3D transducers on the symptomatic side only. Radiographs of the affected wrist were obtained for comparison. US scans were then read by 7 blinded readers to determine accuracy of diagnosis and interrater reliability. An AI network was trained and tested against human readers.
Results: Novice, intermediate and expert human readers had mean 2D/3DUS sensitivities of 0.59, 0.84 and 0.88 respectively and the AI network sensitivity was 0.91.
Conclusions: Study data suggests that 3DUS and portable 2DUS are reliable in detecting DRF. Accuracy of human diagnosis increased with reader experience, and in our data AI diagnosis was comparable to that of expert human readers.