Purpose: to predict vestibular schwannoma (VS) response to radiosurgery by applying machine learning (ML) algorithms on radiomic features extracted from pre-treatment magnetic resonance (MR) images. Methods: patients with VS treated with radiosurgery in two Centers from 2004 to 2016 were retrospectively evaluated. Brain T1-weighted contrast-enhanced MR images were acquired before and at 24 and 36 months after treatment. Clinical and treatment data were collected contextually. Treatment responses were assessed considering the VS volume variation based on pre- and post-radiosurgery MR images at both time points. Tumors were semi-automatically segmented and radiomic features were extracted. Four ML algorithms (Random Forest, Support Vector Machine, Neural Network, and extreme Gradient Boosting) were trained and tested for treatment response (i.e., increased or non-increased tumor volume) using nested cross-validation. For training, feature selection was performed using the Least Absolute Shrinkage and Selection Operator, and the selected features were used as input to separately build the four ML classification algorithms. To overcome class imbalance during training, Synthetic Minority Oversampling Technique was used. Finally, trained models were tested on the corresponding held out set of patients to evaluate balanced accuracy, sensitivity, and specificity. Results: 108 patients treated with Cyberknife^® were retrieved; an increased tumor volume was observed at 24 months in 12 patients, and at 36 months in another group of 12 patients. The Neural Network was the best predictive algorithm for response at 24 (balanced accuracy 73% ± 18%, specificity 85% ± 12%, sensitivity 60% ± 42%) and 36 months (balanced accuracy 65% ± 12%, specificity 83% ± 9%, sensitivity 47% ± 27%). Conclusions: radiomics may predict VS response to radiosurgery avoiding long-term follow-up as well as unnecessary treatment.

Accurate classification of clinically significant prostate cancer remains a major challenge. While multiparametric MRI (mpMRI) has improved lesion detection, effective categorization in accordance to the Prostate Imaging Reporting and Data System (PI-RADS) remains complex. In this study, we propose and evaluate three complementary approaches for automated PI-RADS classification differing in the way in which the features are extracted from the mpMRI imaging sequences. The first approach extracts hand-crafted radiomic features from manually segmented lesions using the PyRadiomics library. The second approach extends this by integrating fully automated lesion and zonal segmentation to simulate a real-world, manual-free pipeline. The third approach utilizes a custom convolutional neural network (CNN) to learn high-level features images and lesion masks directly. The images come from Apparent diffusion coefficient (ADC), diffusion-weighted imaging (DWI), and T2-weighted (T2W) imaging. The features issued by the three methods were used to train a set of machine learning models for multi-class PI-RADS classification, specifically targeting the clinically relevant categories 3, 4, and 5. Results show that ADC-derived features consistently yield superior performance, with one of the ensemble models reaching an AUC of 0.83. Combining features across all sequences further improved robustness (AUC = 0.84). PI-RADS 5 classification was most reliable (AUC ≥ 0.94), whereas PI-RADS 3 remained the most difficult to distinguish. Our findings highlight the effectiveness of ADC features and the advantage of combining automated and deep learning-based strategies for robust prostate cancer risk stratification.