Diagnostic value of artificial intelligence-assisted system for pulmonary nodules: a retrospective observational study

Abstract

Abstract Background: To evaluate the diagnostic value of artificial intelligence-assisted system (AIAS) for pulmonary nodules. Method: This observational study retrospectively enrolled patients with pulmonary nodules with clear pathological findings in the Third Affiliated Hospital of Jinzhou Medical University between July 2019 and May 2021. Receiver operating characteristic curve (ROC) and multivariate logistic regression model were used to evaluate the value of AIAS in the qualitative diagnosis of pulmonary nodules. Result: A total of 112 pulmonary nodules were enrolled in this study, the degree of pulmonary nodules invasion were resulted in benign nodules (35 cases, 31.2%) and malignant nodules (77 cases, 68.8%). There were significant differences between nodules in the benign and malignant groups in terms of age (P = 0.005), average of CT value (P = 0.030), nodule volume (P < 0.001) and malignant signs on the nodule surface (P < 0.001). Multivariate logistic regression analysis showed that the nodule volume (OR = 1.007, 95% CI: 1.003~1.010, P < 0.001) and malignant signs (OR = 7.983, 95% CI: 1.667~38.231, P = 0.009) were independent risk factors for the degree of pulmonary nodules invasion. The sensitivity and specificity of the nodule volume for diagnosing the degree of pulmonary nodules invasion were 83.3% and 88.6% when the nodal volume was 748.98 mm3. Moreover, the malignant signs could diagnose the degree of pulmonary nodules invasion with a sensitivity of 85.7% and a specificity of 71.4%. The nature of the pulmonary nodules identified by AIAS agreed well with the pathological findings by Kappa concordance test (Kappa value = 0.809, P < 0.05). Conclusion: The AIAS may have good accuracy in the qualitative diagnosis of pulmonary nodules and might be used as an auxiliary diagnostic tool for clinicians to distinguish benign and malignant pulmonary nodules. Classification No. of China Map: R734.2