Background: 50-75% of vertebral fractures do not come to medical attention, [1] as they are asymptomatic or there are difficulties in determining the cause of possible symptoms. [2] Improvements in detecting both symptomatic and asymptomatic vertebral compression fractures would increase the potential for treatment of osteoporosis in order to prevent subsequent fractures and improve morbidity and mortality. [2]
Objective: To evaluate the utility of the Zebra automated compression fracture detection algorithm in the identification of vertebral fractures in patients aged 50 years or older having CT chest and/or abdomen scans performed for any reason at a tertiary referral hospital.
Methods: 104 CT abdomen scans and 103 CT chest scans were identified in a retrospective consecutive series. Zebra’s compression fracture detection algorithm was applied to these scans with this process repeated for reproducibility testing. These scans were also assessed by two blinded expert radiologists using predefined criteria to identify and classify vertebral fractures with consensus reached. A search of the original written reports of these scans was undertaken to assess if a compression fracture had been noted in routine reporting.
Results: The sensitivity and specificity of Zebra’s compression fracture detection algorithm will be analysed against the gold standard of expert radiology opinion and compared to that of routine radiology reporting. There will be subgroup analysis of age, gender and reason for request as well as the type of CT scanner used to assess if the automated technology is scanner sensitive.
Conclusion: The utility of Zebra’s automated compression fracture detection algorithm to identify vertebral compression fractures will be analysed in order to further evaluate its use to improve case finding and treatment of osteoporotic vertebral fractures.