VUR grading, VUR + UTI.
| Article DOI | https://doi.org/10.1016/j.eclinm.2024.102466 |
| Objective | To develop and validate a deep-VCUG with ensemble learning for automatic VUR grading from VCUG images |
| AI Approach | CNN, Ensemble learning |
| Data Source(s) | Multi-institutional series (1948 images, from 5 institutions) |
| Model Input | 512 x 512 pixels VCUG images |
| Model Outcome | Unilateral and Bilateral VUR, VUR Grade |
| Model Metrics | Unilateral VUR: AUC of 0.96 (internal), AUC of 0.94 (external) Bilateral VUR: AUC of 0.96 (internal), 0.92 (external) |
| Model Usability | Code is publicly available. Dataset is available upon-request. |
AI = Artificial intelligence, CNN = Convolutional neural network, VCUG = Voiding cystourethrogram, VUR = Vesicoureteral Reflux, AUC = Area under the receiver operator characteristic
AIPEDURO 
Leave a comment