Deep Learning-Based Algorithm Accurately Detects Incidental Pulmonary Embolism on Contrast-Enhanced CT Scans

Strategic Insights -

A recent study conducted by researchers at Emory University has demonstrated the effectiveness of a deep learning-based algorithm in detecting incidental pulmonary embolism (iPE) on contrast-enhanced computed tomography (CT) scans. The algorithm, developed by Avicenna.AI, was able to accurately identify iPE in 87.8% of cases, with a sensitivity of 92.0% and an accuracy of 90.0%. The study also found that the likelihood of iPE increases with increased body CT imaging.

Key Takeaways:

  • The deep learning-based algorithm, CINA-iPE, was able to accurately identify incidental pulmonary embolism (iPE) in 87.8% of cases.
  • The algorithm had a sensitivity of 92.0% and an accuracy of 90.0%.
  • The likelihood of iPE increases with increased body CT imaging.
  • The algorithm correctly identified 159/181 exams positive for PE and 184/200 exams negative for PE.
  • The algorithm missed 22 pulmonary embolisms, with 45.5% of them being complex cases and subject to disagreement among reviewers.
  • Automated results were available to interpreting physicians within 1.5 minutes of data acquisition.
  • The algorithm is useful for increasing the accuracy or speed of detection of iPE.

Statistics:

  • The algorithm had an accuracy of 90.0% (95% CI: 86.6%-92.8%).
  • The sensitivity of the algorithm was 87.8% (95% CI: 82.2%-92.2%).
  • The specificity of the algorithm was 92.0% (95% CI: 87.3%-95.4%).
  • The algorithm correctly identified 159/181 exams positive for PE in 87.8% of cases.
  • The algorithm missed 22 pulmonary embolisms, with 45.5% of them being complex cases and subject to disagreement among reviewers.
  • The time from data acquisition to processing results was 1.5 ± 0.5 (mean ± SD, 95% CI: 1.4%-1.5%) minutes.

Sources:

  • NewsRx. Findings from Emory University Provides New Data about Pulmonary Embolism (Deep learning-based algorithm for automatic detection of incidental pulmonary embolism on contrast-enhanced CT: a multicenter multivendor study). Cardiovascular Week. October 20, 2025; p 195.
  • Deep learning-based algorithm for automatic detection of incidental pulmonary embolism on contrast-enhanced CT: a multicenter multivendor study. Radiology Advances, 2025;2(4).