| ObjectivesBased on the iodine-based images and iodine-suppressed images which are acquired from material decomposition of Spectral CT, this study is aimed to quantitatively analyze the patients with pulmonary nodules or mass, and discuss the value of this technique in the differential diagnosis of lung cancer.MethodsThis was a prospective study of 81 consecutive patients suspected or confirmed of having pulmonary mass or nodules by the previous chest radiography or cross-sectional images. The informed consent was obtained from all patients before they received Spectral CT scan at arterial phase (AP) and venous phase (VP). Finally,70 patients (34 men,36 women; age range,18-80 years; median age,53.9 years) were included, among whom there were 50 patients with primary lung cancers (adenocarcinoma in 31 cases, squamous cell carcinoma in 12 cases, small cell carcinoma in 5 cases, gland spine cell carcinoma in 1 case, mesenchymal malignancies in 1 case) and 20 patients diagnosed as inflammatory mass or nodules. All cases of lung cancer were pathologically diagnosed by aspiration biopsy or surgery; cases of inflammatory mass were pathologically diagnosed by biopsy or clinical follow-up. CT examination were performed on a Discovery CT750HD scanner (GE Healthcare, WI). The routine CT scan was conducted under 120kVp while the enhanced scan was underwent at GSI mode with nonionic contrast agent, which was injected at a dosage standard of 450mg/kg within 30s. The following AP (30s after the injection) and VP GSI scan (60s after the injection) were performed at a tube current of 260mA, a tube rotation rate of 0.6 seconds, a helical pitch of 1.375, a slice thickness of 8mm, an interval of 5mm. Having been reconstructed via adaptive statistical iterative reconstruction technique (ASiR) with a slice thickness of 1.25mm images of routine scan, AP and VP scan were acquired and then analyzed by AW4.6 workstation. Thus, we obtained the iodine-based images (AP and VP)and iodine-suppressed images(AP and VP). Data analyzed is as followed:1.the iodine concentration (IC) of both lung lesions and aorta in the same slice at AP and VP, and the NIC by calculation.2. the CT and SD value of lung lesions in the routine scan images and iodine-suppressed images of AP and VP, and the SNR by calculation.3.The enhanced CT value, which is the difference value of the CT value in AP and VP 70keV images minus the non-enhanced(conventional or virtual) CT value. All the data mentioned above is an average value of three measurement-and-calculation. SPSS 13.0 was used to statistically analyzed the data and test its distribution with homogeneity of variance test. The differences between groups were compared by means of ANOVA(homogeneity of variance) or Tamhane’s T2 test, or by ways of Kruskal-Wallisfen non parametric analysis(non-normal distribution). Using receiver operating characteristic curve (ROC curve) to calculate the dynamic index of iodine concentration, NIC and enhanced CT value(virtual). the best diagnostic point. Receiver operating characteristic (ROC) curve was drawn to calculate the iodine concentration and the optimal diagnostic point of indications, eg. NIC.ResultsComparation of IC among lung caner group, hypervascular inflammatory lesions group and TB groupAt AP, IC in lung cancers were lower than in hypervascular inflammatory occupying lesion(P=0.001, P<0.05), the area under the ROC curve was 0.977, the value of critical diagnosis was 2.0 mg/cm3, sensitivity was 97.6% and specificity was 95.2%; at VP, IC in lung cancers were higher than in TB(P=0.02, P<0.05),the area under the ROC curve was 0.857, the value of critical diagnosis was 1.0 mg/cm3, sensitivity was 90.5% and specificity was 66.7%.Comparation of NIC among lung caner group, hypervascular inflammatory lesions group and TB groupAt VP, NIC in lung cancers were higher than in TB(P=0.003, P<0.05), the area under the ROC curve was 0.941, the value of critical diagnosis was 0.19, sensitivity was 84.6% and specificity was 100%. Comparation of virtual enhanced CT value among lung caner group, hypervascular inflammatory lesions group and TB groupAt AP, the virtual enhanced CT numbers of lung cancers were lower than hypervascular inflammatory lesion group(P=0.001, P<0.05), the area under the ROC curve was 0.774, the value of critical diagnosis was 68HU, sensitivity was 96.4% and specificity was 98%; At VP, the virtual enhanced CT numbers of lung cancer group were lower than hypervascular inflammatory lesion group (P=0.04, P<0.05), the area under the ROC curve was 0.774, the value of critical diagnosis was 58HU, sensitivity was 66.7% and specificity was 75%;the virtual enhanced CT value of lung cancer group were higher than in TB(P=0.01, P<0.05), the area under the ROC curve was 0.945, the value of critical diagnosis was 21HU, sensitivity was 90.5% and specificity was 75.2%.The IC, NIC and virtual enhanced CT value in the adenocarcinoma, squamous cell carcinoma, small cell carcinoma were compared and showed no significant difference during AP and VP (P>0.05).The total effective radiation dose(ED) in conventional non-enhanced scan and dual-phase enhanced scan were 10.71±.30 mSV, the ED in conventional non-enhanced scan was 3.41±0.26 mSV, which could be reduced by VNE, therefore, the total ED in dual-phase enhanced scan was 7.30±0.10 mSV.ConclusionsSpectral CT with gemstone spectral imaging(GSI) is a low-dose method of functional CT imaging, which provides iodine-based images and iodine-suppressed images by material decomposition.The iodine concentration(IC) and normalized iodine concentration (NIC) obtained in iodine-based image could be used as objective and quantitative parameters in valuing pulmonary lesions’ enhancement. Iodine-suppressed image (a type of VNE) could be a proxy of the conventional non-enhanced images for quantitative evaluation of pulmonary lesions and reduce the patient effective radiation dose about 1/3. On balance, Spectral CT with material decomposition could be valuable method for differentiation lung cancer from inflammatory lesions. |
PDF Full Text Request