Table 2.
Details on artificial intelligence algorithm categories and validation strategies used in the papers included in the systematic review
| First author | Algorithm category | Cross-validation | Internal clinical validation | External clinical validation |
|---|---|---|---|---|
| Cheng [19] | Deep learning | N/A | Yes | No |
| Chiu [20] | Deep learning | N/A | Yes | No |
| Cronin [21] | Deep learning | N/A | No | No |
| Cunningham [22] | Deep learning | K-fold cross-validation | No | No |
| Di Cosmo [23]/Smerilli [24] a | Deep learning | K-fold cross-validation | Yes | No |
| Droppelmann [25] | Conventional machine learning | K-fold cross-validation | Yes | No |
| Du Toit [26] | Deep learning | N/A | Yes | No |
| Lee [27] | Deep learning | N/A | Yes | No |
| Lin [28] | Deep learning | N/A | Yes | No |
| Loram [29] | Conventional machine learning, deep learning | Leave-one-out cross-validation | No | No |
| Lyu [30] | Conventional machine learning, deep learning | N/A | Yes | No |
| Marzola [31] | Deep learning | N/A | Yes | No |
| Rosa [32] | Conventional machine learning | N/A | No | No |
| Saleh [33] | Deep learning | K-fold cross-validation | No | No |
| Tiulpin [34] | Conventional machine learning | Leave-one-out cross-validation | No | No |
| Yu [35] | Deep learning | N/A | No | No |