Development of a prediction model for pulmonary nodules using circulating tumor cells combined with the uAI platform

To explore the clinical application value of combining circulating tumor cell (CTC) detection with the artificial intelligence imaging software "uAI platform" in predicting the pathological nature of pulmonary nodules (PN). Develop a joint diagnostic system based on the uAI platform and quantitative detection of CTCs, enable simultaneous classification of pulmonary nodules as benign or malignant and assess the degree of infiltration.

A total of 76 patients with pulmonary nodules undergoing surgical treatment were enrolled. Preoperatively, three-dimensional nodule risk stratification (low、medium、high risk) was performed using the uAI platform, and CTC high-throughput detection was conducted. Key indicators were selected through multi-group comparisons (Benign、Malignant、Invasive subgroups) and logistic regression analysis. A multi-dimensional nomogram model was constructed, and its clinical utility was evaluated using ROC curves and clinical decision curves.

Comparison between benign and malignant pulmonary nodule groups revealed significant differences in the risk stratification of the uAI platform (proportion of high-risk: 75.61% vs 34.29%) and in the median value of CTC quantitative detection (P<0.001). Multivariate logistic regression analysis demonstrated that high-risk classification by uAI and CTC quantitative detection were independent predictors of malignancy in pulmonary nodules (P<0.05). The nomogram model constructed based on these factors exhibited excellent discrimination, and its combined diagnostic performance was significantly better than that of single indicators (AUC=0.805 vs uAI 0.730/CTC 0.743).

The combined uAI-CTC model breaks through the limitations of single-dimension diagnosis, enabling risk stratification of malignant pulmonary nodules and quantitative assessment of infiltration, providing evidence-based support for clinical treatment strategies.