In order to early identify and predict RTLI after IMRT in patients with NPC, we developed and validated a radiomics model based on pretreatment MRI of temporal lobe. The radiomics model demonstrated excellent predictive performance in the validation set (AUC, 0.93; sensitivity, 87%; specificity, 97%). The clinical-radiomics nomogram outperformed clinical nomogram, and showed excellent predictive performance of RTLI in patients within different clinical-pathologic subgroups. These findings suggest that the identified radiomics signature has the potential as a biomarker for risk stratification in RTLI, which may potentially improve the quality of life and prognosis in NPC patients.
This abstract and the presentation materials are available to members only; a login is required.