Endoscopic injection in VUR, VUR + UTI.
| Article DOI | https://doi.org/10.21886/2308-6424-2021-9-2-45-55 |
| Objective | To predict the outcomes of a single endoscopic injection of DxHA for correction of VUR |
| AI Approach | Multilayer ANN (multilayer perceptron) with two hidden layers and a sigmoid function of neuronal activation |
| Data Source(s) | Multi-institutional series (582 children, 783 ureteric units operated on) |
| Model Input | VUR grade, age, sex, presence of ureteral duplication, ureteral dilatation index |
| Model Outcome | To predict successful outcome of DxHA injection treatment |
| Model Metrics | AUC of 0.77, correct prognosis 75%, sensitivity 86%, specificity 65% |
| Model Usability | NA |
AI = Artificial intelligence, ANN = Artificial Neural Network
AIPEDURO 
Leave a comment