Recent research indicates that AI can assist clinicians in accurately identifying vertebral fractures on CT scans, achieving results comparable to experienced specialists. Detecting vertebral fractures is crucial, as they often indicate underlying osteoporosis, which can lead to more fractures. Unfortunately, these injuries are frequently missed during scans due to overlap with other vertebral issues.
In a study involving 3,548 thoracic and lumbar vertebrae, researchers evaluated the performance of eight human raters against four deep learning models and a commercial AI tool. The results showed that AI models matched the diagnostic accuracy of radiology residents in recognizing moderate to severe fractures, signaling potential for enhanced screening in clinical settings.
As AI technology continues advancing, it may offer earlier detection and intervention for osteoporosis, reducing future fracture risks. Further research is necessary, but these findings suggest a promising integration of AI in routine imaging workflows.
Reference: E O Riedel et al. Eur Radiol. 2026; DOI:10.1007/s00330-026-12393-y.
Source link
