Abstract
<jats:p>Background: This study aimed to develop a stratified diagnostic strategy for differentiating focal organizing pneumonia (FOP) from peripheral lung cancer (PLC) by integrating CT-based morphological features with ¹⁸F-FDG PET metabolic parameters. Methods: A retrospective analysis included 102 patients with FOP and 135 patients with PLC confirmed by pathology or follow-up. All patients underwent 18F-FDG PET/CT. Lesions were stratified by diameter and compared between the FOP and PLC groups based on CT morphological features and PET metabolic parameters. A quantitative CT score was developed from significant morphological features using multivariate logistic regression and its diagnostic performance was evaluated. The performance of CT score, PET metabolic parameters and combined models were evaluated using ROC and DeLong test analysis. Results: In lesions≤10 mm, CT density distribution was the key discriminator. For 10-30 mm lesions, gender, multiple CT morphological features and metabolic tumor volume (MTV) showed significant differences (all P&lt; 0.05). The CT scoring system achieved excellent discriminatory performance (AUC=0.961), performing better than PET (AUC=0.662, P&lt; 0.001) and comparably to the combined PET/CT model (AUC=0.963, P=0.48). In lesions&gt;30 mm, the combined PET/CT model (AUC=0.901) outperformed the CT score alone (AUC=0.824, P=0.03). Subgroup analysis revealed that PET added additional diagnostic value for lesions with CT scores &lt; 49.7, increasing the AUC from 0.646 to 0.892 (P=0.02); however, the benefit was minimal for lesions with CT scores≥49.7. Conclusions: We propose a lesion size-guided diagnostic strategy integrating CT scoring and 18F-FDG PET. For small lesions (≤10 mm), CT follow-up is sufficient. For medium-sized lesions (10-30 mm), CT scoring alone provides high diagnostic accuracy. For larger lesions (&gt;30 mm), CT scoring serves as initial screening, with PET reserved for low-score cases. This integrated diagnostic strategy can reduce unnecessary PET scans while preserving diagnostic efficacy, offering significant clinical value.</jats:p>