能谱与常规模式下减影去骨法脑CTA图像质量的对比研究
发布时间：2019-06-20 04:12
【摘要】：目的：探讨能谱CTA脑动脉与脑组织最佳对比噪声比曲线峰值的变化规律,并以此为基础前瞻性对比评价能谱模式扫描与常规模式扫描减影去骨法脑动脉CTA的图像质量。 材料和方法：连续74名接受头颅CTA检查的患者随机分配至能谱扫描组和常规扫描组,采用GE DISCOVERY CT750HD扫描仪分别以能谱模式及常规模式进行扫描,观察能谱模式下虚拟单能量未减影增强脑动脉与脑实质最佳对比噪声比曲线,记录其两个峰值对应的虚拟单能量水平keV1、keV2,比较能谱扫描模式(?)得到的三组图像(140kVp、keV1、keV2)以及常规扫描模式下的一组图像(routine,120kVp)减影去骨法CT图像的脑动脉CT值、对比度、背景噪声、对比噪声比、信噪比以及脑动脉CTA VR+segment/autoselect法重建图像的去骨效果和脑动脉图像质量(均为5分法评价,1-差、2-较差、3-中等、4-良好、5-优秀) 结果：虚拟单能量未减影增强脑动脉与脑实质最佳对比噪声比曲线均表现为驼峰状双峰曲线,两个峰值对应的keV1、keV2的单能量水平分别为(55.19±2.05)keV、(64.92±1.16)keV。减影后140kVp、keV1、keV2、routine四组图像间颅骨去除效果没有差别χ2=3.85,p=0.280.05)。两位放射科医生对脑血管图像质量的主观评价结果一致性较好,Kappa值为0.65,u=12.53,p=0.00(0.05)。Ridit分析R1140kVP=0.33, R1keV1=0.70, R1keV2=0.55, R1routine=0.42, χ2=35.52, P=0.000.05;R2,40kVP=0.31, R2keV1=0.68, R2keV2=0.57, R2routine=0.44, χ2=33.52, P=0.00(0.05四组图像的主观评分明显有差别,四组图像质量从好到差排序为keV1、keV2、 routine、140kVp, keV1组图像明显好于其它三组。减影后140kVp、keV1、keV2、 routine四组图像间的脑动脉CT值、对比度、背景噪声、对比噪声比、信噪比均有差别,F值分别为38.849、38.620、7.995、11.342、10.982(p=0.000.05)。四组图像中,keV1组脑动脉CT值、对比度最高,keV2组与routine组居中,140kVp组最低；背景噪声keV2组最低；对比噪声比及信噪比keV1和keV2组无差别,140kVp与routine组无差别,keVl、keV2组高于140kVp和routine两组。 结论： 1.未减影能谱扫描模式增强后脑动脉与脑组织最佳对比噪声比双峰曲线峰值对应的keV1组的平均虚拟单能量值约为(55.19±2.05)keV,keV2组的平均虚拟单能量值为(64.92±1.16)keV。 2.两种模式扫描四组图像用减影以及VR+segment/autoselect软件综合法去骨的效果相仿并都可基本消除减影后的残留颅骨,优于现行的去骨法。 3.与常规扫描模式减影脑CTA相比,能谱扫描模式下取得的140kVp图像,其减影CTA图像质量差,因此只适合于用于作为血管强化程度的参考图像。 4.能谱扫描模式的keVl、keV2组减影去骨法脑CTA可以明显提高图像质量,更清晰的显示脑动脉,可以取代常规扫描模式更好的显示脑动脉。keV1组图像较keV2组图像质量更优秀,keV1组的平均值约为55keV,因此建议以55keV为能谱扫描常规后处理的虚拟单能量图像显示脑动脉,从而优化能谱数据的后处理流程。
[Abstract]:Objective: to investigate the changes of the peak value of the optimal contrast noise ratio curve between cerebral artery and brain tissue in energy spectrum CTA, and to evaluate the image quality of CTA by energy spectrum mode scanning and conventional mode scanning subtraction. Materials and methods: 74 consecutive patients undergoing skull CTA were randomly assigned to energy spectrum scanning group and routine scanning group. GE DISCOVERY CT750HD scanner was used to scan in energy spectrum mode and conventional mode respectively. The optimal contrast noise ratio curve of virtual single energy unsubtraction enhanced cerebral artery and brain parenchyma in energy spectrum mode was observed, and the virtual single energy level keV1,keV2, comparative energy spectrum scanning mode corresponding to the two peaks was recorded. The CT values, contrast, background noise, contrast noise, signal-to-noise ratio, signal-to-noise ratio (SNR), bone removal effect and cerebral artery image quality of three groups of images (140kVP, KeV1, Kev2) and a group of images (routine,120kVp) subtracted from conventional scanning mode (CT) subtractive images, contrast, background noise, signal-to-noise ratio, signal-to-noise ratio (SNR) and cerebral artery image quality (all evaluated by 5 points method, 1-difference, 2-poor, 3-medium, 4-good. Results: the optimal contrast noise ratio curve between cerebral artery and cerebral parenchyma was hump bimodal curve, and the single energy level of keV1,keV2 corresponding to the two peaks was (5519 卤2.05) keV, (64.92 卤1.16) keV., respectively. After subtraction, there was no difference in skull removal effect among the four groups of images, such as 140kVp, KeV1, KeV2 and routine. There was no difference in skull removal effect between the four groups (P 鈮，
本文编号：2502913
[Abstract]:Objective: to investigate the changes of the peak value of the optimal contrast noise ratio curve between cerebral artery and brain tissue in energy spectrum CTA, and to evaluate the image quality of CTA by energy spectrum mode scanning and conventional mode scanning subtraction. Materials and methods: 74 consecutive patients undergoing skull CTA were randomly assigned to energy spectrum scanning group and routine scanning group. GE DISCOVERY CT750HD scanner was used to scan in energy spectrum mode and conventional mode respectively. The optimal contrast noise ratio curve of virtual single energy unsubtraction enhanced cerebral artery and brain parenchyma in energy spectrum mode was observed, and the virtual single energy level keV1,keV2, comparative energy spectrum scanning mode corresponding to the two peaks was recorded. The CT values, contrast, background noise, contrast noise, signal-to-noise ratio, signal-to-noise ratio (SNR), bone removal effect and cerebral artery image quality of three groups of images (140kVP, KeV1, Kev2) and a group of images (routine,120kVp) subtracted from conventional scanning mode (CT) subtractive images, contrast, background noise, signal-to-noise ratio, signal-to-noise ratio (SNR) and cerebral artery image quality (all evaluated by 5 points method, 1-difference, 2-poor, 3-medium, 4-good. Results: the optimal contrast noise ratio curve between cerebral artery and cerebral parenchyma was hump bimodal curve, and the single energy level of keV1,keV2 corresponding to the two peaks was (5519 卤2.05) keV, (64.92 卤1.16) keV., respectively. After subtraction, there was no difference in skull removal effect among the four groups of images, such as 140kVp, KeV1, KeV2 and routine. There was no difference in skull removal effect between the four groups (P 鈮，
本文编号：2502913
本文链接：https://www.wllwen.com/yixuelunwen/yundongyixue/2502913.html
最近更新
教材专著
