Abstract
Abstract
Tumor heterogeneity plays a central role in treatment resistance, disease progression, and diagnostic uncertainty. However, it may be overlooked by traditional 2D histology. Accurate 3D assessment of tumor microarchitecture is therefore important for capturing its spatial complexity. Micro-CT, a well-established imaging modality now emerging for high-resolution 3D virtual histology of soft tissues, provides a promising alternative. Combined with radiomics, this technique enables interpretable, quantitative characterization of tumor biology beyond visual inspection. In this study, we analyzed radiomics signatures of a large cohort of thyroid tumors (418 patients) using micro-CT imaging of tissue microarrays. We achieved robust classification of (i) neoplastic versus non-neoplastic thyroid tissues, (ii) papillary thyroid carcinoma versus follicular thyroid neoplasm, and (iii) BRAF V600E mutation status. Shapley additive explanations were used to reveal key visual traits driving these classification decisions. Exploratory analysis in a limited TERT cohort (8 mutated vs 103 wild-type) identified prospective radiomics patterns associated with TERT promoter mutations, suggesting potential surrogate imaging biomarkers that warrant further investigation. Micro-CT radiomics shows promise as a complementary tool for diagnostic classification in thyroid cancer and offers a platform for quantitative 3D tissue characterization pending broader validation.