AI-Powered Fracture Detection Solution Reduces Risk of Misdiagnosis

By MedImaging International staff writers
Posted on 03 Jul 2023

BoneView integrates seamlessly with all IT infrastructures (Photo courtesy of GLEAMER)

The analysis of radiographs has become increasingly challenging due to escalating workloads in emergency and radiology departments. Missed fractures account for up to 80% of diagnostic errors in emergency departments, as a result of diagnostic discrepancies between initial radiographic interpretations and final reads by board-certified radiologists. This often leads to care delays and unnecessary harm to patients. However, with artificial intelligence (AI) assistance, both sensitivity and potential specificity of fracture detection can be improved by radiologists and non-radiologists without extending reading time. Now, an AI-enabled fracture detection solution offers an instant and automatic second read of trauma X-rays, fully integrated into the reading workflow.

GLEAMER’s (Paris, France) BoneView is the first AI-powered solution to be approved for detecting fractures in both adult and pediatric populations. BoneView employs advanced AI algorithms to support various healthcare professionals, including radiologists, orthopedic surgeons, emergency physicians, rheumatologists, family physicians, and physician assistants. It aids in diagnosing fractures promptly and accurately, thereby improving patient outcomes and reducing misdiagnosis risks. BoneView's AI algorithms can identify subtle abnormalities, enhancing diagnostic sensitivity, specificity, and accuracy. Moreover, the software integrates effortlessly into any workflow, enhancing efficiency and saving time for radiologists.

Reading performances of both radiologists and emergency physicians with and without BoneView's assistance have been evaluated on multi-centric data, including imaging equipment not present in the training set. The results indicated statistically significant improvements. BoneView has shown to reduce false-negative rates (missed fractures) by 30%, while also decreasing reading time. Furthermore, a recent study endorsed BoneView's effectiveness on US pediatric data, demonstrating a sensitivity of 91% for children and teenagers. The FDA has granted BoneView clearance for use in children over two years old, indicating a substantial advancement in improving fracture detection and diagnosis through AI. BoneView analyzes over seven million images across the world every year and is deployed in more than 22 countries across Europe, the Middle East, Asia-Pacific, and North America. Over 5000 radiologists and emergency physicians depend on BoneView in their clinical routines.