Response to chemotherapy in Wilms tumor
| Article DOI | https://doi.org/10.4111/icu.20240135 |
| Objective | To predict the response to preoperative chemotherapy in Wilms tumor using contrast-enhanced computed tomography (CT) |
| AI Approach | SVM, decision tree, KNN, LR, multi-layer perceptron |
| Data Source(s) | Single-centre institutional series (54 patients, 63 tumors with pre-/post-chemotherapy CT) |
| Model Input | Integrated features from pre-therapy CT: shape (sphericity, elongation, spherical harmonics), functionality-based (enhancement slopes across phases), texture from multiple phases |
| Model Outcome | Response to chemotherapy (binary) |
| Model Metrics | SVM accuracy 95%, sensitivity 96%, specificity 94%, accuracy 95%, F1 0.97 |
| Model Usability | Code not provided; dataset not publicly available. |
AI = Artificial intelligence, SVM = Support vector machines, KNN = K-nearest neighbors, AUC = Area under the receiver operator characteristic
AIPEDURO 
Leave a comment