Enhancing Identification of High-Risk cN0 Lung Adenocarcinoma Patients Using MRI-Based Radiomic Features

Harim Kim et al. developed an MRI-based radiomics model to predict high-risk pathologic features in clinical N0 lung adenocarcinoma patients. Using AVIEW software, they extracted radiomics features from T1- and T2-weighted MRI images of 72 patients selected based on CT and PET/CT imaging criteria. Their model predicted features like micropapillary/solid patterns (MPsol), spread through air spaces (STAS), and poor differentiation with high accuracy (AUC values: 0.751-0.907). These findings correlated with poorer survival and increased nodal upstaging. AVIEW facilitated precise tumor ROI delineation and data analysis, ensuring robust imaging-pathology correlation. The study demonstrated MRI's potential in identifying high-grade lung cancer features preoperatively, supporting more precise treatment planning. Limitations included a small sample size and lack of external validation. This research paves the way for integrating advanced imaging and radiomics in personalizing lung cancer management by accurately characterizing histologic risk features and optimizing surgical decisions.