Key points are not available for this paper at this time.
Objective: To assess the predictive value of radiomics from the 2-cm edema zone surrounding the postoperative residual cavity for progression-free survival (PFS), using habitat analysis based on multimodal magnetic resonance imaging (MRI) and integrating clinical data to construct a nomogram model. Methods: This retrospective study analyzed MRI and clinical data from 89 postoperative glioma patients. The 2-cm edema zone surrounding the postoperative residual cavity was defined as the region of interest (ROI), and habitat subregions were created using K-means clustering based on contrast-enhanced T1-weighted imaging (CE-T1WI) and apparent diffusion coefficient (ADC) sequences. Radiomic features were extracted from the ROI and each habitat subregion, followed by Least Absolute Shrinkage and Selection Operator (LASSO)-Cox selection to generate radiomic scores. Clinical, traditional radiomic, and high-risk habitat models were constructed, and the high-risk habitat nomogram was further developed and evaluated. Results: Four habitat subregions were identified. A total of 944 radiomic features were extracted from each subregion and the ROI; the most relevant features were used to generate radiomic scores. The high-risk habitat nomogram was constructed by combining clinical factors. The nomogram showed good calibration, with observed values closely matching predictions. In the validation cohort, the time-dependent AUCs for predicting 1-, 2-, and 3-year PFS were 0.813, 0.933, and 0.930, respectively. Compared with the clinical and traditional radiomic models, the high-risk habitat nomogram achieved a C-index of 0.916. Conclusion: The nomogram based on high-risk habitats in the 2-cm edema zone surrounding the postoperative residual cavity provides significant predictive value for PFS and aids in targeting postoperative radiotherapy.
Cheng et al. (Tue,) studied this question.