TABLE 7.
Overview of the predictive models in radiomic studies
| Classification | Prognostics | Total | ||
|---|---|---|---|---|
| Single model | Regression | 40 | 11 | 51 |
| Cox | 0 | 16 | 16 | |
| Random Forest | 9 | 3 | 12 | |
| Support Vector Machine | 8 | 2 | 10 | |
| Statistics | 3 | 3 | 6 | |
| Multiple models | Support Vector Machine | 8 | 11 | 19 |
| Regression | 6 | 8 | 14 | |
| Naïve Bayes | 6 | 5 | 11 | |
| Discriminant Analysis | 7 | 2 | 9 | |
| k‐Nearest Neighbor | 4 | 4 | 8 | |
| Random Forest | 8 | 0 | 8 | |
| Decision Tree | 1 | 5 | 6 | |
| Neural Network | 3 | 3 | 6 | |
| Boosting | 0 | 5 | 5 | |
| Cox | 0 | 5 | 5 | |
The studies were separated by single model and multiple models. Regression, primarily logistic regression, was most commonly used in the reviewed studies.