Table 5.
The diagnostic performance of artificial intelligence in coronary stenosis.
| Study | Year | Methods | Sensitivity | specificity | PPV | NPV | Accuracy |
|---|---|---|---|---|---|---|---|
| Kang et al. (33) | 2015 | SVM | 93% | 95% | NA | NA | 94% |
| Chen et al. (59) | 2020 | DL | 94% | 63% | 94% | 59% | NA |
| Arnoldi et al. (60) | 2010 | Computer-aided | 100% | 65% | 58% | 100% | 100% |
| Kelm et al. (61) | 2011 | Supervised Learning | 97.62% | 67.14% | NA | 99.77% | NA |
| Goldenberg et al. (62) | 2012 | CAST | >90% | 40%−70% | NA | > 95% | NA |
DL, deep learning; SVM, support vector machine; CAD, coronary artery disease; PPV, positive predictive value; NPV, negative predictive value; QCA, quantitative coronary angiography; CCTA, coronary CT angiography; CAST, computer-aided simple triage; NA, not applicable.