Vesicoureteral Reflux and Urinary Tract Infection
| Article DOI | https://doi.org/10.1097/JU.0000000000000186 |
| Objective | To predict the probability of recurrent UTI and associated VUR after initial UTI, but before VCUG |
| AI Approach | Optimal classification trees, random forest, gradient-boosted trees |
| Data Source(s) | Multi-institutional trial dataset (RIVUR, 305 patients; CUTIE, 195 patients) |
| Model Input | Age, age, gender, race, weight, antibiotic resistance in urine culture, urine protein, dysuria, medications, antibiotics in last 6 months, blood pressure |
| Model Outcome | VUR-associated recurrent UTI |
| Model Metrics | AUROC = 0.76 |
| Model Usability | Easily accessible decision trees, freely accessible GitHub and available mobile application (Appendix 5) |
AIPEDURO 
Leave a comment