Application of artificial intelligence-assisted diagnosis and treatment system in the diagnosis of pulmonary nodules: a clinical analysis of 1650 cases
10.3760/cma.j.cn112434-20221220-00389
- VernacularTitle:人工智能辅助诊疗系统在肺小结节诊断中的应用——附1 650例临床分析
- Author:
Zheng WANG
1
;
Hong ZHANG
;
Jun LV
;
Xiangfeng LUO
;
Daqiang SUN
Author Information
1. 天津大学附属天津市胸科医院胸外科，天津 300222
- Keywords:
Artificial intelligence;
Pulmonary nodules;
Auxiliary diagnosis
- From:
Chinese Journal of Thoracic and Cardiovascular Surgery
2023;39(8):466-471
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To evaluate the effectiveness of the artificial intelligence-assisted diagnosis and treatment system in distinguishing benign and malignant lung nodules.Methods:Clinical data of 1 650 patients with pulmonary nodules admitted to the Tianjin Chest Hospital Affiliated to Tianjin University from January 2015 to October 2022 were retrospectively analyzed, which included a total of 1 650 nodules. The malignant probability was compared to and analyzedpredict the benign and malignant nodules.Results:A total of 1 650 nodules were included, which were divided into a malignant tumor group(n=1 391, 84.3%) and a benign lesion group(n=259, 15.7%). Between the malignant tumor group and the benign lesion group, the malignant probability was significantly different, and the malignant probability could better distinguish malignant nodules and benign lesions[(78.85±22.91)% vs.(54.91±28.68)%, P<0.001]. The area under the curve( AUC) was 0.768. The critical value of malignant probability for diagnosis of lung cancer was 81.3% with a sensitivity of 0.620 and specificity of 0.815. Stratified analysis results showed that the accuracy of the AI intelligent system for diagnosing pulmonary nodules with the sizes of 0-1 cm, 1-2 cm and 2-3 cm was also increased, and the areas under ROC curve were 0.717, 0.769 and 0.804, respectively. Conclusion:The malignant probability of lung nodules calculated by the artificial intelligence-assisted diagnosis and treatment system can be used in the assessment of the preoperative benign and malignant lung nodules.
