Radiomic Feature Stability in Low-Dose CT and 3D Segmentation for Pulmonary Nodules: Phantom and Multicenter Evaluation

This is the authors' abstract. We don't add key points for this paper.

Artificial intelligence–based three-dimensional region of interest segmentation and low-dose CT imaging were used to identify a stable radiomic feature subset for pulmonary nodules, which demonstrated robust diagnostic performance across 1098 clinical ground-glass nodules (AUC, 0.86–0.90).