Circulating tumor cell combined with artificial intelligence to establish a model for diagnosing the malignancy of pulmonary nodules

Abstract

Background Exploring the clinical application value of combining circulating tumor cell (CTC) with artificial intelligence in predicting the pathological nature of pulmonary nodules. Constructing a prediction model based on factors related to lung cancer to provide reliable prediction criteria for clinical doctors to predict the pathological nature of pulmonary nodules, in order to guide clinical doctors in judging the benign and malignant nature and infiltration degree of pulmonary nodules (PN). Methods This study included a total of 76 patients with PN who underwent surgical treatment. Based on preoperative imaging of the patients, an artificial intelligence imaging system called "United Imaging Intelligence" was used to classify the pulmonary nodules into three levels of "low risk", "medium risk", and "high risk", and the preoperative CTC level of the patients was recorded. Multiple logistic regression analysis was used to analyze the risk factors affecting the nature of the PN and to construct relevant column charts. Receiver operating characteristic (ROC) curves were used to analyze the diagnostic value of artificial intelligence and CTC levels for the nature of PN lesions. Results The artificial intelligence model for grouping benign and malignant PN and the difference in CTC levels have statistical significance (P < 0.05). The results of multifactor logistic regression analysis showed that artificial intelligence high-risk grouping, CTC level, and age are independent risk factors affecting the nature of PN (P < 0.05). We also constructed a column chart to guide clinical doctors in treatment. The area under the curve (AUC) for the artificial intelligence risk grouping and CTC level diagnosis of malignant PN were 78.9% and 74.3%, respectively. Conclusion Artificial intelligence model combined with CTC detection helps improve the accuracy of lung nodule characterization diagnosis and assists in guiding clinical decisions.