Objective: To investigate the application of experience-based and Magnetic Resonance Imaging (MRI) radiomics in differentiating benign and malignant pulmonary nodules and masses. Methods: Sixty patients with pulmonary nodules or masses admitted from April 2024 to April 2025 were selected as study subjects. Surgical pathology was used as the gold standard to explore the differential diagnostic value of experience-based and MRI radiomics. Results: Twenty one MRI radiomic features were selected for screening, with 14 features removed and 7 remaining for normality testing. Using surgical pathology as the gold standard, among the 60 patients, 36 were benign and 24 were malignant, with detection rates of 60.00% and 40.00%, respectively. The detection rates of lobulation, vascular convergence sign, and GGO components in benign pulmonary nodules/masses were lower than those in malignant ones (P < 0.05). The radiomics formula was Radscore = Intercept + Weight (Feature) × Feature, with the calculated formula being Radscore = 0.16 - 0.32 × Sphericity + 0.11 × Mass. The AUC of the UTE Rad-score model was 0.672, which was lower than the AUC of the UTE nomogram (0.789) and the combined model (0.832) (P < 0.05). Conclusion: Experience-based and MRI radiomics can play a significant role in differentiating benign and malignant pulmonary nodules and masses, with the combined model demonstrating more prominent diagnostic value.