Glioblastoma multiforme (GBM), the most aggressive primary brain tumor, is characterized by rapid recurrence and poor prognosis despite multimodal therapy. Accurate differentiation of GBM recurrence from treatment-related effects (TrE), including pseudoprogression (PsP) and radiation necrosis (RN), remains a major clinical challenge due to overlapping imaging features on conventional Magnetic Resonance Imaging (MRI). Radiomics has emerged as a noninvasive quantitative imaging approach that extracts high-dimensional features from medical images and integrates them with machine learning algorithms. This review summarizes recent advances in radiomics for recurrent GBM, including characterization of recurrence patterns, preoperative and postoperative recurrence risk prediction, spatial localization of recurrent lesions, and differentiation of recurrence from TrE. Key technical and clinical challenges are also discussed, including data heterogeneity, limited external validation, model generalizability, and biological interpretability. By linking imaging biomarkers with clinical and biological insights, radiomics demonstrates significant translational potential for improving recurrence assessment in GBM. Future efforts should focus on multicenter validation, standardized imaging protocols, and enhanced interpretability to facilitate reliable clinical implementation.
Zhang et al. (Tue,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: