双层探测器光谱CT在肺原发恶性肿瘤、肺慢性炎症及肺结核鉴别诊断中的应用价值

Clinical value of dual-layer spectral detector CT in distinguishing diagnosis of pulmonary primary malignant tumor,chronic inflammation and tuberculosis

目的探讨双层探测器光谱CT(DLCT)参数和常规CT征象在肺原发恶性肿瘤、肺慢性炎症及肺结核鉴别诊断中的应用价值。方法选取2020年8月-2021年6月在甘肃省人民医院行DLCT胸部增强扫描并获得病理结果的345例肺部疾病患者,按照病理结果分为肺原发恶性肿瘤组(n=187)、肺慢性炎症组(n=101)与肺结核组(n=57)。回顾性分析三组病变的常规CT征象并测量其DLCT参数,对差异有统计学意义的参数进行logistic回归分析,分别建立常规CT征象诊断模型、DLCT参数诊断模型及联合诊断模型,使用受试者工作特征曲线(ROC)分析各模型的诊断效能,采用Delong检验比较各模型的曲线下面积(AUC)。结果三组病变的常规CT征象包括病灶距胸膜距离(P=0.009)、形态(P<0.001)、密度(P=0.001)、与肺组织分界(P=0.001),以及有无分叶(P<0.001)、液化坏死(P=0.003)、血管集束征(P<0.001)、晕征(P=0.003)、卫星灶(P=0.045)、胸腔积液(P=0.002)、纵隔淋巴结大(P<0.001),DLCT参数包括动脉期(AP)及静脉期(VP)有效原子序数(Z_(eff))、碘密度(IC)、标准化碘密度(NIC)、能谱曲线斜率(λ_(HU))、动脉强化分数(AEF)(P<0.001)均存在明显差异。在肺原发恶性肿瘤和慢性炎症的鉴别诊断中,病灶与肺组织分界(P=0.009)、分叶(P<0.001)、液化坏死(P<0.001)、晕征(P=0.025)、纵隔淋巴结大(P<0.001)、λ_(HU)-AP(P=0.037)和λ_(HU)-VP(P=0.029)为独立影响因素;在原发恶性肿瘤和肺结核的鉴别诊断中,病灶形态(P=0.019)、血管集束征(P=0.009)、卫星灶(P=0.006)、胸腔积液(P=0.001)、AEF(P=0.041)、λ_(HU)-AP(P=0.038)及λ_(HU)-VP(P<0.001)为独立影响因素;在慢性炎症和肺结核的鉴别诊断中,胸腔积液(P=0.002)、纵隔淋巴结大(P<0.001)、NIC-VP(P=0.001)、Z_(eff)-VP(P=0.043)、λ_(HU)-AP(P=0.015)及λ_(HU)-VP(P=0.023)为独立影响因素。以病理结果为金标准,鉴别肺原发恶性肿瘤与肺慢性炎症、肺原发恶性肿瘤与肺结核、肺慢性炎症与肺结核时,常规CT征象诊断模型的AUC分别为0.827、0.770及0.753,DLCT参数诊断模型的AUC分别为0.905、0.909及0.824,常规CT征象和DLCT参数联合诊断模型的AUC分别为0.929、0.942、0.889。结论DLCT参数联合常规CT征象可提高肺原发恶性肿瘤、肺慢性炎症及肺结核的鉴别诊断效能。

Objective To explore the clinical value of dual-layer spectral detector CT(DLCT)in distinguishing diagnosis of pulmonary primary malignant tumor,chronic inflammation and tuberculosis by measuring and analyzing the parameters and conventional CT signs.Methods The clinical data of 345 patients with pulmonary lesions were collected from August 2020 to June 2021,who underwent DLCT chest enhanced scan and obtained pathological results in People's Hospital of Gansu Province,and then divided into three groups:pulmonary primary malignant tumor group(n=187),chronic inflammation group(n=101)and tuberculosis group(n=57).The conventional CT signs of the three groups were retrospectively analyzed and the DLCT parameters were measured.The logistic regression analysis was performed for parameters with statistically significant differences,and then conventional CT signs diagnostic model,DLCT parameter diagnostic model and combined diagnostic model were established.The receiver operating characteristic(ROC)curve was used to analyze the diagnostic efficacy of each model.Delong test was used to compare the AUC of each models.Results In distinguishing the conventional CT signs of the three lesions,statistical differences existed in the following indicators:the distance of lesions to pleura(P=0.009),morphology(P<0.001),density(P=0.001),the boundary between lesions and lung(P=0.001),lobulation(P<0.001),liquefaction necrosis(P=0.003),vascular cluster sign(P<0.001),halo sign(P=0.003),satellite focus(P=0.045),pleural effusion(P=0.002),enlarged lymph nodes in the mediastinum(P<0.001),effective atomic number(Zeff),iodine concentration(IC),normalization iodine concentration(NIC),energy spectrum curve slope(λ_(HU)),and arterial enhancement fraction(AEF)(P<0.001)both in the arterial phase(AP)and venous phase(VP).In the differential diagnosis of pulmonary primary malignant tumor and chronic inflammation,the boundary between lesion and lung tissue(P=0.009),lobulation(P<0.001),liquefaction necrosis(P<0.001),halo sign(P=0.025),mediastinal lymphadenopathy(P<0.001),λ_(HU)-AP(P=0.037)andλ_(HU)-VP(P=0.029)are independent influencing factors.In the differential diagnosis of pulmonary primary malignant tumor and tuberculosis,lesion morphology(P=0.019),vascular cluster sign(P=0.009),satellite focus(P=0.006),pleural effusion(P=0.001),AEF(P=0.041),λ_(HU)-AP(P=0.038)andλ_(HU)-VP(P<0.001)are independent influencing factors.Pleural effusion(P=0.002),mediastinal lymphadenopathy(P<0.001),NIC-VP(P=0.001),Zeff-VP(P=0.043),λ_(HU)-AP(P=0.015)andλ_(HU)-VP(P=0.023)are independent influencing factors in the differential diagnosis of chronic inflammation and tuberculosis.To pathology results for the gold standard,the AUC of conventional CT signs diagnostic model for distinguishing pulmonary primary malignant tumor and chronic inflammation,pulmonary primary malignant tumor and tuberculosis,chronic inflammation and tuberculosis were 0.827,0.770 and 0.753.The AUC of DLCT parameter values for distinguishing pulmonary primary malignant tumor,chronic inflammation and tuberculosis were 0.9050.909 and 0.824.The AUC of the combined model for distinguishing pulmonary primary malignant tumor,chronic inflammation and tuberculosis were 0.929,0.942 and 0.889.Conclusion DLCT parameters combined with conventional CT signs may improve the differential diagnosis efficiency of pulmonary primary malignant tumor,chronic inflammation and tuberculosis.