<p>Summary of the synthetic training dataset.</p>

Duplicate/overlap, identifier-only, and permuted-outcome controls assessed leakage.</p> <p>Results</p><p>Pooled OOF AUROC was 0.955443 for CatBoost, 0.955324 for LightGBM, 0.955066 for HistGradientBoosting, and 0.952874 for logistic regression. Corresponding average precision ranged from 0.945912 to 0.948828 and Brier scores from 0.081116 to 0.083518. Calibration intercepts ranged from −0.001299 to 0.002153 and slopes from 0.998896 to 1.004388.

Mean optimism was 0.000017–0.002686. No duplicated identifiers, duplicated feature profiles, or train–test feature-profile overlap were found; identifier-only AUROC was 0.500121 and mean permuted-outcome AUROC was 0.500615.</p> <p>Conclusions</p><p>All four models discriminated strongly and showed close internal calibration, while gradient-boosting models had only marginally higher AUROC than logistic regression. The principal contribution is a fully rerunnable, controlled benchmark rather than a novel algorithm or a clinically validated prediction model.

External performance in real clinical cohorts remains unknown.</p></div>