MRI-based Radiomics for Prognosis of Non-Midline, Pediatric High-Grade Gliomas: analysis code

Analysis code accompanying the study "MRI-based Radiomics for Prognosis of Non-Midline, Pediatric High-Grade Gliomas" (Journal of Neuro-Oncology, 2026). Reproduces the radiomics survival-analysis pipeline: minimum-redundancy maximum-relevance (mRMR) feature selection within k-fold cross-validation, multivariable Cox proportional-hazards modeling of overall survival, risk-score Kaplan-Meier stratification, and the anaplastic astrocytoma vs. glioblastoma comparison. Version 1.1.0 adds a bring-your-own-data toolkit: the frozen published Cox model (model/paper_model.json), an external-validation script (validate.py: concordance, Kaplan-Meier risk stratification, calibration slope), a training script for new cohorts (train.py: full mRMR+Cox pipeline or coefficient re-estimation), the exact PyRadiomics 2.2.0 extraction settings (pyradiomics_params.yaml), and a step-by-step guide for outside institutions.