| Autor: |
Chieregato, Matteo, Spatola, Giorgio, Morassi, Mauro, Migliorati, Karol, Cobelli, Milena, Bassetti, Chiara, Bagnalasta, Matteo, Bnà, Claudio, Galelli, Marco, Franzin, Alberto |
| Předmět: |
|
| Zdroj: |
Journal of Radiosurgery & SBRT; 2022 Supplement, Vol. 8, p180-180, 1p |
| Abstrakt: |
Meningiomas are benign lesions that can spread all over the skull. Tumors adjacent to a sinus or draining vein can result in venous congestion and in formation of peritumoral edema. Sterotactic radiosurgery can worsen an existing edema or make a new edema appear in a delayed fashion (5-10% of cases). In the present work we built a predictive machine learning model for new edema formation following meningioma gamma knife radiosurgery. Our dataset consists of single-institution patients treated for meningioma with Gamma Knife radiosurgery, either in single fraction or in 3-5 fractions, with at least six months imaging follow-up. The predictive model is built on radiomic features from planning MRI, patients clinical characteristics and plan dosimetric data. Machine learning methods are used to tackle data imbalance and multiple meningiomas treatment in single patient. Game theoretical explanations and counterfactual examples are utilized to provide insights at individual prediction level. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
