Table 3.
Univariate and multivariate meta-regression analyses for identifying covariates to explain heterogeneity among studies on clinical, hormonal (laboratory) and imaging data-based ML models for the diagnosis of CPP.
| Covariates | Multivariate meta-regression | SEN and 95% CI | SPE and 95% CI | ||
|---|---|---|---|---|---|
| LR (Chi-square test) |
P | I2 index (%) | |||
|
Feature
Image feature (n =4) Non-image feature (n = 2) |
6.13 | 0.04 | 67 |
0.90 [0.83 - 0.97] 0.53 [0.25 - 0.81] |
0.90 [0.86 - 0.94] 0.83 [0.77 - 0.89] |
|
Classifier
LR (n =4) XGBoost (n = 4) |
4.22 | 0.02 | 53 |
0.93 [0.87 - 0.99] 0.62 [0.42 - 0.82] |
0.88 [0.84 - 0.93] 0.82 [0.73 - 0.90] |
|
Classifier
RF (n =5) LR (n = 4) |
4.68 | 0.01 | 57 |
0.91 [0.87 - 0.95] 0.77 [0.47 - 1.00] |
0.91 [0.87 - 0.95] 0.81 [0.77 - 0.86] |
|
Classifier
GBM (n =1) RF (n = 5) |
2.99 | 0.22 | 33 | – | – |
LR, logistic regression; XGBoost, extreme gradient boosting; RF, random forest; GBM, gradient boosting machine; SPE, Specificity; SEN, Sensitivity. P value of <0.05 was considered statistically significant.