Glioblastoma (GBM) is an aggressive brain tumor that often infiltrates beyond its visible boundaries, making treatment challenging with standard surgical and chemoradiotherapy approaches.
A new method was developed that combines expert insights and data augmentation to improve predictions of tumor infiltration using preoperative magnetic resonance imaging (mpMRI) scans from 229 patients.
The model was validated through cross-institutional tests, showing varying effectiveness in predicting tumor recurrence, with odds ratios indicating strong potential for guiding targeted treatment strategies.