Predictive value of quantitative artificial intelligence parameters combined with circulating tumor cell detection in lung adenocarcinoma infiltration

Abstract

Abstract Objective Investigation of the value of artificial intelligence (AI) quantitative parameters combined with circulating tumor cell testing (CTC) in lung adenocarcinoma infiltration.Methods Images of 127 surgically confirmed samples of lung adenocarcinomas were analyzed retrospectively between January 2020 and December 2022, and based on postoperative pathology, the lung adenocarcinomas were divided into the non-infiltrating group (65 cases) and the infiltrating group (62 cases), with the latter, including ICA. Five sets of quantitative indices, namely, the longest diameter, volume, mass, mean CT value, and maximum CT value, of each nodule, were analyzed using AI analysis software. The patients were subjected to the CTC detection test prior to the surgery.The differences in the above five quantitative indices between the two groups were determined, following which the ROC curve analysis and calculations for the area under the curve (AUC), 95% CI, sensitivity, specificity, critical value, and compliance rate were performed for each group. The subsequent multifactorial binary logistic regression analysis of each quantitative parameter and CTC revealed volume, mean CT, and CTC as the independent risk factors for pulmonary nodule infiltration prediction. A multifactorial logistic regression analysis was then performed to construct a combined model (for volume, mean CT, and CTC). The diagnostic efficacy of the combined model was compared based on the volume, mean CT, and CTC, respectively, using ROC curves and Z-tests.Results The quantitative indices (longest diameter, volume, mass, maximum CT and mean CT of each nodule) were higher in the infiltration group compared to the non-infiltration group, and the difference between the two groups was statistically significant (P < 0.05). The number of CTC-positive cases was higher in the infiltration group compared to the non-infiltration group, and the difference was statistically significant (P < 0.05).The area under the curve (AUC) for the longest diameter, volume, mass, maximum CT, and mean CT was 0.845, 0.850, 0.756, 0.727, and 0.871, respectively. The highest sensitivity for each quantitative parameter was obtained as 88.7% for CTC, the highest specificity was 93.8% for maximum CT, and the highest compliance was 83.5% for maximum CT.The volume, mean CT, and CTC were revealed as independent risk factors for predicting the infiltrative nature of pulmonary nodules, with the respective odds ratio (OR) of 1.001, 1.006, and 5.065; the corresponding 95% CI were 1.000–1.001, 1.002–1.009, and 1.269–20.210, respectively, with P < 0.05.The mean value of the AUC of the combined model was 0.934, with 95% CI in the range of 0.887 to 0.982, a sensitivity of 91.9%, a specificity of 87.7%, and a compliance rate of 88.20%. The diagnostic efficacy of the combined model was significantly higher than the independent use of the volume, mean CT, and CTC parameters for prediction (Z = 2.315, 2.290, and 4.7, respectively, all P-values were < 0.05).Conclusion The quantitative AI parameters of volume and mean CT value combined with CTC provide a better preoperative prediction of the infiltrative nature of lung adenocarcinoma.