Machine Learning Model to Predict Postmastectomy Breast Reconstruction Complications - PubMed
3 hours ago
- #breast reconstruction
- #complication prediction
- #machine learning
- Researchers developed and validated machine learning models to predict major complications after postmastectomy breast reconstruction.
- The study used retrospective data from 411 female patients at two U.S. academic centers, focusing on implant-based or autologous reconstruction.
- An extreme gradient boosting model outperformed a random forest model, achieving an AUROC of 0.83 and identifying top predictors like smoking and radiotherapy.
- The model's performance remained consistent across different reconstructive modalities, supporting personalized risk assessment for shared decision-making.