St. Luke's University Health Network

Emergency department clinicians needed faster and more accurate interpretation of X-ray images to identify bone trauma such as fractures, dislocations, effusions, and lesions. Subtle findings like buckle fractures are easy to miss under the time pressure of an ER setting, and radiologists serve as a final review layer — meaning earlier triage accuracy directly impacts patient outcomes.

St. Luke's deployed Gleamer BoneView AI — trained on a massive database of X-ray images using deep learning algorithms — across all acute care hospital ERs and Care Now walk-in urgent care centers. The system analyzes X-ray images in real time, placing a yellow bounding box around areas of concern before images are forwarded to radiologists for final review. It is applied on average to more than 1,000 X-ray images every day network-wide.

Clinical studies show the tool reduces missed fractures by up to 30%. The deployment is now active at all St. Luke's hospital emergency rooms following the final installation at the Grand View campus in Sellersville. The network is involved in over 100 AI projects, one-third of which have been fully integrated into operational workflows, with Gleamer BoneView AI representing the first network-wide medical imaging AI initiative.

• Deploying AI as a "second set of eyes" before radiologist review accelerates diagnostic speed without removing human oversight.
• Network-wide rollout requires phased campus-by-campus implementation — St. Luke's began with select ERs before expanding to all locations.
• AI diagnostic tools in residency training programs create compounding value by building physician fluency with next-generation imaging workflows.