Cureus. 2026 Jul 2;18(7):e111959. doi: 10.7759/cureus.111959. eCollection 2026 Jul.
ABSTRACT
Background MRI radiomics has recently gained attention for evaluating intratumoral heterogeneity in thoracic malignancies. Thymic epithelial tumors (TETs) include a wide spectrum from the low-risk thymoma group (LRG) to thymic carcinoma (TC), but imaging-based differentiation remains challenging. This study aimed to assess the utility of MRI texture analysis for differentiating TET subtypes according to the 2021 WHO classification. Methods Forty-seven patients with histologically proven TETs (19 LRG, 18 high-risk thymoma group (HRG), and 10 TC) underwent preoperative MRI including T1-, T2-, diffusion-weighted (DWI), and apparent diffusion coefficient (ADC) sequences. Texture analysis was performed using LIFEx software (version 7.6.0; French Alternative Energies and Atomic Energy Commission, Paris, France). A total of 41 texture features were extracted from manually delineated three-dimensional tumor volumes. Statistical comparisons were performed using Kruskal-Wallis and Mann-Whitney U tests with Benjamini-Hochberg false discovery rate correction. Diagnostic performance was evaluated using receiver operating characteristic (ROC) analysis. Results Among the 47 lesions, 24, 28, and 37 texture parameters differed significantly between LRG-HRG, LRG-TC, and HRG-TC, respectively. The highest area under the curve (AUC) values were 0.785 for DWI coefficient of variation in differentiating LRG from HRG, 0.933 for T1WI morphological radius in differentiating LRG from TC, and 0.867 for ADC intensity variance in differentiating HRG from TC. Conclusions MRI-based texture analysis provides a promising noninvasive approach for differentiating thymic epithelial tumor subtypes according to the WHO classification. Sequence-specific texture parameters demonstrated moderate to high diagnostic performance, particularly for differentiating low-risk thymoma from thymic carcinoma. These findings suggest that MRI radiomics may contribute to preoperative risk stratification and treatment planning in patients with thymic epithelial tumors.
PMID:42542899 | PMC:PMC13428655 | DOI:10.7759/cureus.111959
