An international trial has confirmed the strong performance of AI software for opportunistically detecting osteopenia and osteoporosis from chest radiographs.
In an ongoing study consisting of over 1,000 consecutive adult patients from the U.S., India, and Brazil who had received both poster-anterior chest x-rays and dual-energy x-ray absorptiometry (DEXA) exams, presenter Anushree Burade, MBBS, of Massachusetts General Hospital and Harvard Medical School, and colleagues found that an AI tool yielded generalizable findings for differentiating between the three classes of bone mineral density (BMD).
For distinguishing between osteoporosis and non-osteoporosis patient groups, the model provided a receiver operating characteristic area under the curve (ROC AUC) of 0.836, sensitivity of 68.8%, positive predictive value of 23.6%, negative predictive value of 97.7%, and accuracy of 70.5%. It also had an ROC AUC of 0.719 for distinguishing osteopenia from normal groups.
No significant differences were found between the three countries.
“Our ongoing study supports the generalizability of CXR-based BMD prediction AI model across data from the three countries with high AUC and NPV,” the authors wrote. “Given the global underutilization of dedicated osteoporosis screening with DEXA, such AI-based opportunistic screening using CXRs may help bridge the diagnostic gap.”
Learn about all of the latest findings from this study by attending this Sunday morning talk.
