能谱CT定量分析在鉴别诊断肺鳞癌与腺癌中的应用价值

Value of Spectral CT Imaging in Differentiating Plumonary Squamous Cell Carcinoma and Adenocarcinoma

目的探讨能谱CT在肺鳞癌与腺癌鉴别诊断中的应用价值。方法选择2011年8月至2012年6月在我院行能谱CT扫描并经病理证实的肺癌患者46例(鳞癌28例,腺癌18例),对比分析两组病例平扫及增强扫描后的能谱特征参数及能谱曲线。结果平扫鳞癌的有效原子序数及钙含量均大于腺癌,分别为(7.82±0.08)和(7.73±0.07)、(4.69±1.27)和(3.01±1.53)mg.mL-1,差异有统计学意义(P均<0.05)。增强扫描70keV单能量下鳞癌和腺癌的强化程度差异无统计学意义(t=1.26,P=0.21);鳞癌能谱曲线斜率及标准碘含量(1.90±0.22、0.14±0.02)大于腺癌(1.55±0.13、0.12±0.01),P均<0.05。ROC曲线分析显示曲线斜率鉴别鳞癌和腺癌具有较高的灵敏度(86%)和特异度(89%),且曲线下面积0.93。结论肺鳞癌和腺癌具有不同的能谱曲线和能谱特征参数,能谱CT可为鳞癌和腺癌的鉴别诊断提供一种新的多参数的方法。

Objective To investigate the feasibility of differentiating ptumonary squamous adenocarcinoma by using dual energy spectral CT imaging. Methods 46 lung cancer patients （28 patients with squamous cell carcinoma and 18 patients with adenocarcinoma） underwent plain and enhanced CT scans with spectral imaging mode. Spectral CT imaging related parameters for the cancers such as calcium concentra-tion and effective - Z in the plain CT and normalized iodine concentration （ NIC, to that of aorta） and spectral HU curve in the enhanced CT were obtained. The increment of CT value in 70 keV was calculated as △CT = CTenhanced-CTplain. The slope k of spectral HU curve was calculated as k = [ HU （40keV） - HU （90keV） ]/50. All results were compared with independent samples t test. Receiver operating characteristic （ROC） curves were generated to determine parameters for optimizing sensitivity and specificity in differentiating two different lung cancers. Results ① In plain CT： The effective -Z （7.82 ± 0.08） and calcium concentration （4.69 ± 1.27mg·mL^-1） of squamous cell carcinoma were significantly higher than those of adenocarcinoma [ （ 7.73 ±0. 07 ）, （3.01 ±_1.53 mg · mL^-1） ] （ t = 4.06,3.70, P 〈 0.05 ）. ②In enhanced CT： There was no difference in increment of CT value in 70 keV between squamous cell carcinoma and adenocarcinoma （t = 1.26, P =0.21 ）. The slope k and NIC were （ 1.90 ±0.22, 0.14 ±0.02） and （ 1.55 ±0.13,0.12 ±0.01 ） for squamous cell carcinoma and adenocarcinoma, respectively, the differences in the slope k and NIC between two pathological types of lung cancers showed all significant （t = 6.81,2.81, P 〈 0.05 ）. ③The slope in the enhanced CT provided high sensitivity and specificity in differentiating squamous cell carcinoma and adenoear-einoma （86% and 89% , respectively） ,and the area under the ROC curve was 0.93. Conclusion Spectral CT imaging provided additional parameters than conventional CT for plumonary squamous cell carcinoma and adenocarcinoma that were statistically different, and it may become an effective diagnostic method to improve the sensitivity and specificity in differentiating these two pathological types of lung cancers.