Differentiating peripheral lung cancers from inflammatory masses using dual energy spectral CT imaging / 中华放射学杂志

ABSTRACT

Objectives To investigate the clinical significance of dual energy spectral CT (DESCT) in quantitatively differentiating peripheral lung cancers from pulmonary inflammatory masses.Methods Sixty patients with 35 lung cancers and 25 inflammatory masses underwent DESCT to get arterial phase (AP) images and venous phase (VP) images.Iodine concentrations in the central and peripheral zone of the masses were measured and normalized to the aorta as normalised iodine concentration (NIC).The difference of NIC between central and peripheral zone of the masses (dNIC) was calculated.The spectral attenuation curve was obtained automatically and the slope of curve (λHU) was also calculated in the two groups.The quantitative parameters was presented as M (Q1,Q3),and Wilcoxon signed rank test was used to compare above two independent samples.Receiver operating characteristic (ROC) curves were generated to calculate the sensitivity and specificity.Results NICs in the central zone of peripheral lung cancers were significantly lower than that of inflammatory massesmean NICs were 0.03 (0,0.05) versus 0.12 (0.07,0.20) in AP,and 0.14 (0.12,0.19) versus 0.30 (0.21,0.57) in VP (Z=-4.14,-3.70,respectively,P＜0.01).While the dNIC values of lung cancers were significantly higher than that of inflammatory massesdNIC values were 0.08 (0.05,0.11) versus 0.04 (-0.02,0.08) in AP,and 0.23 (0.17,0.34)versus 0.07 (-0.04,0.08) in VP(Z=-2.56,-4.00,respectively,P＜0.05).Mean λHU values of lung cancers were also lower than inflammatory masses1.03 (0.67,1.67)versus 2.75 (1.61,3.19) in AP,and 1.58 (1.30,2.17) versus 3.25 (2.37,4.54) in VP (Z=-3.90,-4.42 respectively,P＜0.01).According to ROC curves,cutoff value of λHU =2.11 in VP had the highest sensitivity (89％) and specificity (91％) in differentiating peripheral lung cancers from inflammatory masses.Conclusions Contrast-enhanced dual energy spectral CT imaging with some quantitative parameters such as normalised iodine concentration,dNIC,and the slope of spectral attenuation curves may be a promising new method for differentiating peripheral lung cancers from inflammatory masses.