Table 3. Performances of physicians without and with support of AI in classification of moderate-severe vs. non-moderate-severe in MR patients.
| Physician group | Sensitivity (%) (95% CI) | Specificity (%) (95% CI) | |||
|---|---|---|---|---|---|
| Without AI | With AI | Without AI | With AI | ||
| All physicians (n=9) | 77.0 (70.9–82.1) | 86.7 (80.3–91.2) | 91.5 (87.8–94.1) | 90.5 (86.7–93.2) | |
| Junior physicians (n=3) | 84.0 (72.6–91.3) | 95.8 (94.3–96.9) | 87.4 (78.9–92.8) | 84.4 (82.2–86.4) | |
| Intermediate physicians (n=3) | 77.5 (68.7–84.3) | 85.4 (76.1–91.6) | 93.1 (89.7–95.4) | 91.3 (89.3–93.0) | |
| Senior physicians (n=3) | 69.5 (68.7–70.2) | 78.9 (74.7–82.5) | 93.9 (92.4–95.2) | 95.7 (93.4–97.2) | |
AI, artificial intelligence; MR, mitral regurgitation.