双能CT单能谱图像及非线性融合技术对孤立性肺结节应用价值研究

发布时间：2018-04-30 01:10

本文选题：双能CT + 单能谱图像　；参考：《重庆医科大学》2017年硕士论文

【摘要】：第一部分双能CT单能谱成像检测不同性质孤立性肺结节的实验研究目的:分析单能谱图像与孤立性肺结节检出率及其图像质量的关系。方法:采用男性胸部仿真体模,并随机置入不同密度(-800HU、-650HU与100HU)及大小(3mm、5mm、8mm、10mm、12mm)的球形模拟肺结节,使用西门子双源炫速CT(SOMATOM definition flash)双能量模式(80/140KV)对体模进行扫描,提取出不同能量水平的单能谱图像后对每组图像进行图像质量及检出率分析。分组1:每种密度肺结节中各包含5种不同直径肺结节各6枚,共计90枚;分组2:选取-650HU磨玻璃肺结节,每种直径肺结节各9枚,共计45枚。使用SPSS软件进行统计分析,肺结节的噪声、SNR、CNR采用ANOVA方差分析,若差异有统计学意义,则进一步进行两两对比(若方差齐则使用LSD检验,方差不齐使用Dunnett T3检验);肺结节检出率的比较采用Fisher确切概率法。结果:-800HU、-650HU与100HU的模拟肺结节在不同单能谱图像上噪声、SNR及CNR的差异有统计学意义(P0.05),且通过绘制折线图发现在70ke V图像上,各密度肺结节的噪声低而SNR、CNR最高;以70ke V为对照组,65ke V、75ke V与70ke V两两比较,相同密度肺结节图像噪声、SNR及CNR的差异无统计学意义(P0.05);在-650HU肺结节中,5种不同直径的肺结节在每个单能量水平图像上的噪声、SNR及CNR的差异均无统计学意义(P0.05)。在60~140ke V图像上-800HU、-650及100HU肺结节检出率为100%,大于5mm的模拟肺结节在所有单能谱图像上均检出。在40和50ke V图像上-800HU及3mm模拟肺结节检出率低于60~140ke V(P0.05)。结论:胸部双能CT孤立性肺结节成像时,70ke V时图像的噪声最小,其SNR与CNR最高;选择60ke V以上单能谱图像可提高孤立性肺结节的检出率。第二部分双能量CT非线性图像融合技术对孤立性肺结节成像的可行性研究目的:探讨双能量CT非线性融合技术对孤立性肺结节成像的可行性,优化各密度肺结节非线性融合参数。方法:采用男性胸部仿真体模,利用非离子型对比剂碘普罗胺注射液与生理盐水混合模拟增强扫描后不同密度肺结节并置入胸部体模,分别使用西门子双源炫速CT(SOMATOM definition flash)双能模式(80/140k V)及单能120k V(两者CTDI值一致)对体模进行扫描后重建出线性及非线性融合图像后对图像进行图像质量分析。结果:对于各不同CT值的模拟肺结节,BC为0HU,BW为0~30HU时可获得较高的CNR。不同CT值的模拟肺结节在非线性融合、线性融合及120k V图像上的噪声值及SNR差异有统计学意义(P0.05),CNR差异无统计学意义(P0.05)。其中各不同密度模拟肺结节在非线性融合优化图像的SNR显著高于线性融合组及120k V组(P0.05);而图像的噪声在非线性融合优化组低于线性融合组(P0.05)。结论:双能量CT非线性融合优化技术使CT增强扫描后肺结节的SNR明显提高,可为更准确诊断孤立性肺结节提供帮助。
[Abstract]:The first part of the experimental study on the detection of solitary pulmonary nodules with dual energy CT single energy spectrum imaging objective: to analyze the relationship between single energy spectrum images and the detection rate and image quality of solitary pulmonary nodules. Methods: the male chest phantom was used, and the spherical simulated pulmonary nodules with different densities of Hu-800HU-650HU and 100HUU) and size of 3mm / 5mm / 8mm / 10mm / 12mm) were randomly placed. The models were scanned by Siemens dual-source CT(SOMATOM definition flash-mode (80 / 140KV). After extracting single energy spectrum images with different energy levels, the image quality and detection rate of each group of images were analyzed. Group 1: there were 6 pulmonary nodules with 5 different diameters in each density and 90 nodules in different diameters. Group 2 was divided into two groups: -650 Hu milled glass-shaped pulmonary nodules, each with 9 pulmonary nodules, with a total of 45. SPSS software was used for statistical analysis. The noise of pulmonary nodules was analyzed by ANOVA ANOVA. If the difference was statistically significant, further comparison was made (if the variance was homogeneous, LSD test was used. Dunnett T 3 test was used for variance heterogeneity, and Fisher exact probability method was used to compare the detection rate of pulmonary nodules. Results there was significant difference in noise SNR and CNR between the simulated pulmonary nodule between the two groups on different single energy spectrum images (P0.05A). It was found that on the 70ke V image, the noisiness of each density pulmonary nodule was lower and the SNRN was the highest. 70ke V was compared with 70ke V in the control group. There was no significant difference in noise-SNR and CNR in the same density pulmonary nodule images (P 0.05), but there was no significant difference in the noise of 5 lung nodules with different diameters on each single energy level image (P 0.05). On 60~140ke V images, the detectable rate of pulmonary nodules of -800 HU-650 and 100HU was 100. The simulated pulmonary nodules larger than 5mm were detected on all single energy spectrum images. The detectable rate of -800 Hu and 3mm simulated pulmonary nodules on 40 and 50ke V images was lower than that on 60~140ke V V (P 0.05). Conclusion: the SNR and CNR of solitary pulmonary nodules were the least and the SNR and CNR were the highest at 70ke V, and the detection rate of solitary pulmonary nodules could be improved by selecting single energy spectrum images above 60ke V. The second part: feasibility study of dual energy CT nonlinear image fusion technique for solitary pulmonary nodule imaging objective: to explore the feasibility of dual energy CT nonlinear fusion technique in solitary pulmonary nodule imaging. The nonlinear fusion parameters of lung nodules with different densities were optimized. Methods: male chest phantom was used to simulate the pulmonary nodules of different density after contrast injection of non-ionic contrast agent iopramide and normal saline. Siemens dual-source CT(SOMATOM definition flash-two-energy mode (80 / 140kV) and single energy 120kV (the same value of CTDI) were used to reconstruct linear and nonlinear fusion images and image quality analysis. Results: for the simulated pulmonary nodules with different CT values, when BC was 0 HUU BW was 0~30HU, higher CNRs could be obtained. There were significant differences in the noise value and SNR of the simulated pulmonary nodules with different CT values in nonlinear fusion, linear fusion and 120kV images. The SNR of different density simulated pulmonary nodules in nonlinear fusion optimization image is significantly higher than that in linear fusion group and 120kV group, while the noise of image is lower in nonlinear fusion optimization group than that in linear fusion group. Conclusion: dual energy CT nonlinear fusion optimization technique can improve the SNR of pulmonary nodules after enhanced CT scan, which can help to diagnose solitary pulmonary nodules more accurately.
【学位授予单位】：重庆医科大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R734.2;R730.44

【参考文献】