Integrating radiomics and immunology: Non-invasive assessment of intratumoral CD8T cell levels in murine models.

Abstract

Radiomics has emerged as a promising tool for extracting quantitative features from medical images, enabling deeper insights into tumor phenotypes and microenvironmental heterogeneity. Nonetheless, the clinical application of these features is impeded by their inherent non-biological interpretability. It is, therefore, critically important to unravel the biological meaning of these radiomic features. This study aims to address this gap by integrating radiomics with immunology to establish a direct link between imaging biomarkers and immune dynamics. Using murine cancer models, we manipulated the tumor microenvironment with monoclonal antibodies to modulate CD8T cell levels, a critical determinant of antitumoral immunity. Tumors were imaged, segmented, and analyzed for radiomic feature extraction, followed by immune cell quantification. Through exploratory factor analysis, we identified a radiomic signature that reflects intratumoral CD8T cell levels, establishing a novel non-invasive biomarker. This work enhances the explainability of radiomic features and supports their potential role in immune profiling within preclinical models. Future validation in human imaging datasets is necessary to determine their translational relevance for precision oncology.