Skip to main content

Table 3 Results of the controlled experiment in which clinicians diagnosed 300 test images

From: Artificial intelligence improves the accuracy of residents in the diagnosis of hip fractures: a multicenter study

  Accuracy (%) Sensitivity (%) Specificity (%)
With CAD Without CAD P value With CAD Without CAD P value With CAD Without CAD P value
First-year residents (95%CI) 82.1 (78.6, 85.7) 91.2 (89.3, 93.0) < 0.001* 81.9 (79.4, 84.3) 90.5 (88.5, 92.5) < 0.001* 86.5 (83.0, 90.1) 93.2 (91.3, 95.1) 0.08
Second-year residents (95%CI) 85.9 (83.9,87.9) 90.9 (89.3, 92.6) < 0.001* 84.2 (81.4, 87.0) 90.6 (88.0, 93.2) < 0.001* 89.7 (88.2, 91.2) 93.5 (92.4, 94.7) < 0.001*
  1. 95%CI 95% confident intervals
  2. * P < 0.05