Abstract:
Objective To explore the application value of artificial intelligence (AI) combined with multi-slice spiral CT in screening pulmonary nodules in health examination population in government departments.
Methods A retrospective analysis was conducted on the chest CT screening data of health examination population in government departments who had at least diameter of one nodule ≥3 mm. All data were analyzed using AI technology combined with manual film reading. The detection rate of pulmonary nodules was recorded. Based on the risk values predicted by AI technology, the nodules were divided into different risk groups, and the basic characteristics of the nodules were analyzed and compared.
Results The detection rate of pulmonary nodules was 60.4%. High-risk patients were more likely to present as pure ground glass nodules or mixed ground glass nodules. The average age of individuals with pulmonary nodules was higher than those without(P<0.05). There were significant differences in the basic characteristics of pulmonary nodules, including nature, morphology, internal signs, and external signs among different risk groups (P<0.05).
Conclusion Multi-slice spiral CT examination of the chest scan is helpful for screening pulmonary nodules in large-scale physical examination populations. The combination of AI technology and manual film reading can improve the accuracy of pulmonary nodule screening.
下载: