Funding outcome

One of the most common reasons a child in Alberta seeks medical care is an upper extremity injury. Falling on an outstretched hand during activity risks a fracture at the wrist or elbow, or a shoulder rotator cuff tear. In emergency settings, these injuries are usually examined using x-rays which might require travel to a tertiary clinic, time-consuming transfers within an ER and a minor radiation risk. Point of care ultrasound (POCUS) is a faster, radiation-free alternative for detecting all these injuries. The AI model can assist healthcare professionals in making early and accurate diagnoses of wrist and elbow fractures. Unlike x-rays that require transfer to a separate room within ED, POCUS can be performed by nurses at the triage station. We expect this workflow to significantly reduce wait-times, lead to faster identification of fractures, quicker decision-making, and more personalized patient care. We will train AI models to perform key aspects of image analysis including video summarization, segmentation, classification, and domain adaptation. The AI model will be trained to provide realtime feedback on image quality there-by enabling lighlty trained users like triage nurses to perform this examination. This can be especially beneficial in remote and rural regions where specialized expertise is scarce. The development of such AI models can contribute to research in pediatric musculoskeletal health. Implementing an AI model for fracture detection in ultrasound images could contribute to cost-effective and faster triage strategies for pediatric fractures that would reduce ED wait-times and provide more personalized care.