基于肺部多模态图像诊断恶性肺结节的影响因素

Influencing factors of malignant pulmonary nodules based on lung multimodal images

目的探讨多模态影像参数在判断肺结节良、恶性中的应用价值。方法收集来自首都医科大学宣武医院及北京市肿瘤防治研究所肺占位患者病例326例,包括115例良性病例,211例恶性病例。记录多模态影像指标,包括结节大小、最大标准摄取值(maximum standard uptake value, SUVmax)、磨玻璃样变、边缘光滑、有晕征、分叶征、毛刺征、钙化共计8个指标,通过多因素Logistic回归分析筛选出基于肺部多模态图像的恶性肺结节的影响因素。结果多因素Logistic回归分析显示,结节大小、SUVmax、分叶征、毛刺征为恶性肺结节的危险因素,相应的OR值分别为1.31(1.03~1.65)、1.10(1.03~1.18)、7.27(3.57~14.82)、3.16(1.47~6.83)。结节钙化为恶性肺结节的保护因素,OR值为0.13(0.05~0.32)。结论结节大、SUVmax值大、存在分叶征、存在毛刺征、结节无钙化提示肺结节为恶性结节的可能性较大。

Objective To explore the value of multimodal imaging parameters in determining benign and malignant pulmonary nodules.Methods This study collected pulmonary occupying patients from Xuanwu Hospital of Capital Medical University and Beijing Institute of Cancer Prevention and Treatment.A total of 326 cases were collected in this study, including 115 benign cases and 211 malignant cases.Multi-modal imaging indicators were recorded, including nodule size, maximum standard uptake value(SUVmax),ground glass nodules, margin, halo, lobular, spicule, and calcification.Multivariate Logistic regression analysis was used to screen out the influencing factors of malignant pulmonary nodules based on multi-modal images.Results Multivariate Logistic regression analysis showed that nodule size, SUVmax, lobular and spicule were risk factors for malignant pulmonary nodules, and the corresponding OR values were 1.31(1.03-1.65),1.10(1.03-1.18),7.27(3.57-14.82),3.16(1.47-6.83).Calcification was a protective factor for malignant pulmonary nodules, with an OR value of 0.13(0.05-0.32).Conclusions Large nodules, large SUVmax values, lobular signs, burr signs, and no calcification of the nodules suggest that lung nodules are more likely to be malignant nodules.