Diagnostic Value of AI Technology Combined with MSCT in Invasive Lesions of Ground-glass Nodules in Lung Adenocarcinoma
Objective To investigate the diagnostic value of artificial intelligence(AI)technology combined with multi-slice spiral CT(MSCT)in invasive lesions of ground-glass nodules(GGN)in lung adenocarcinoma.Methods The clinical data of 80 patients with GGN(80 nodules)in lung adenocarcinoma who were treated in the hospital from June 2018 to June 2023 were analyzed retrospectively.With pathological results as the gold standard,the diagnostic efficiencies of AI technology,MSCT and their combination in invasive lesions of GGN in lung adenocarcinoma were evaluated.According to examination results,the patients were divided into invasive group and non-invasive group.AI and CT parameters in both groups were analyzed.Results Pathological results showed that there were 38 patients with invasive lesions and 42 patients with non-invasive lesions.The sensitivity,specificity,accuracy,negative and positive predictive values of AI technology or MSCT for diagnosing invasive lesions of GGN in lung adenocarcinoma were lower than those of AI technology combined with MSCT(92.11％,97.62％,95.00％,97.22％and 93.18％).The long diameter,short diameter,maximum CT value,maximum area and volume in the invasive group were larger than those in the non-invasive group,and the minimum CT value was lower than that in the non-invasive group(P＜0.05).The proportions of spicule or lobulation signs on the edge,quasi-circular lesions,vessel convergence signs and pleural indentation signs in the invasive group were higher than those in the non-invasive group(P＜0.05).Conclusion The combination of AI technology and MSCT can effectively improve the diagnostic efficiency of invasive lesions of GGN in lung adenocarcinoma.Large lesion area and volume,large long and short diameters,high maximum CT value,low minimum CT value,spicule or lobulation signs on the edge and irregular shape are important features that can help to diagnose invasive lesions.
万方数据
维普
