能谱CT虚拟平扫图像在腮腺肿瘤检出及显示中的可行性

Feasibility of Energy Spectral CT Virtual Non-contrast Images in the Detection and Display of Parotid Tumors

目的比较能谱CT虚拟平扫图像(VU)、水基图(WB)及常规平扫(CU)对腮腺肿瘤的检出及显示,探讨能谱CT虚拟平扫替代常规平扫诊断腮腺肿瘤的可行性。资料与方法回顾性分析行能谱CT平扫加增强扫描的45例腮腺肿瘤患者的资料,均进行能谱扫描,重建VU及WB图像。测量3组图像各部位的CT值及皮下脂肪噪声,计算腮腺病灶的对比噪声比(CNR),进行主观评分,并记录有效辐射剂量(ED),比较不同图像间主观评分、病灶清晰度显示评分、腮腺肿瘤病灶的CNR及常规平扫加双期增强扫描图像与双期直接增强扫描的ED。结果 CU与VU图像间肌肉、淋巴结的CT值及腮腺肿瘤的相对强化差异无统计学意义(P>0.05)。VU图像噪声高于CU,差异有统计学意义(P=0.005)。WB图像的CNR高于VU图像,VU图像的CNR高于CU图像,差异有统计学意义(P<0.01)。CU、VU和WB图像对腮腺肿瘤病灶的检出率分别为94.02%(63/67)、92.53%(62/67)和89.55%(60/67),差异无统计学意义(P>0.05)。CU与VU的图像质量评分、病灶清晰度评分差异均无统计学意义(P>0.05),均高于WB图像(P均<0.01)。常规平扫加双期增强扫描图像和双期直接增强扫描的ED分别为(3.96±0.14)mSV、(2.49±0.11)mSV,ED降低33.31%(1.47/3.96)。结论能谱CT虚拟平扫可以为腮腺肿瘤的诊断提供一个可靠的CT基线值,替代常规CT平扫后可明显降低辐射剂量。

Purpose To compare energy spectral CT virtual unenhanced (VU) images, water based (WB) image and conventional unenhanced (CU) CT scan for the detection and display of parotid tumors, and to explore the feasibility of using energy spectral CT virtual non-contrast images instead of conventional non-contrast CT scan for the diagnosis of parotid tumors. Materials and Methods A retrospective analysis of 45 patients with parotid tumors who underwent energy spectral CT non-contract scan combined with enhancement scan was performed. All patients received energy spectral scan to reconstruct VU and WB images. The CT values and subcutaneous fat noise of each part of images of the three groups were measured, and the contrast-to-noise ratio (CNR) of the parotid lesions was calculated for subjective scoring and the effective dose (ED) was recorded. The subjective scoring, lesion resolution, CNR of parotid tumor lesions were compared among different images. The comparison between conventional non-contrast scan plus dual-phase contrast enhanced scan images and dual-phase direct contrast enhanced scan was performed. Results The difference in CT value of muscle and lymph nodes as well as the relative enhancement value of parotid tumors between the CU and VU images suggested no statistical significance ( P >0.05). The image noise of VU was higher than that of CU, the difference of which was statistically significant ( P =0.005). The CNR of the WB image was higher than that of the VU image, and the CNR of the VU image was higher than that of the CU image, the difference of which was statistically significant ( P <0.01). The detection rates of CU, VU and WB images for parotid tumor lesions were 94.02%(63/67), 92.53%(62/67) and 89.55%(60/67), respectively;the difference of which was not statistically significant ( P >0.05). There was no statistical significance concerning the difference in scoring of image quality and lesion resolution between CU and VU ( P >0.05), with both higher than WB images ( P <0.01). The ED of conventional non-contrast scan plus dual-phase contrast enhanced scan images and dual-phase direct contrast enhanced scan were (3.96±0.14) mSV and (2.49±0.11) mSV, respectively;and the ED reduction rate was 33.31%(1.47/3.96). Conclusion The energy spectral CT virtual non-contrast scan can provide a reliable CT baseline value for the diagnosis of parotid tumors, which can significantly reduce the radiation dose after replacing the conventional CT non-contrast scan.