Wilms tumor stage, Oncology.
| Article DOI | https://doi.org/10.3389/fped.2022.873035 |
| Objective | To preoperatively predict stages (stage I and non-stage I) Wilms tumor in pediatric patients |
| AI Approach | Support Vector Machines |
| Data Source(s) | Single institutional series (118 patients) |
| Model Input | Pre-op portal venous-phase CT images (15 features: 2 first-order features, 13 texture features) |
| Model Outcome | Wilms Tumor stage (stage I vs other) |
| Model Metrics | Accuracy of 75-79%, AUC of 0.79-0.81 |
| Model Usability | Dataset and code available as supplementary material upon-request. |
AI = Artificial intelligence
AIPEDURO 
Leave a comment