基于多模型的迭代重建技术对低剂量CT结肠成像图像质量的影响

目的 评价Revolution CT应用不同水平基于多模型的迭代重建算法（ASIR-V）对提高低剂量CT结肠成像图像质量的能力。方法 选用离体猪结肠获得模拟息肉30个，使用Revolution CT在不同扫描条件（管电压120 kVp，管电流10、30、50、70、90、100、120、140、160、180、200、220、240、260 mA）下扫描，分别应用6种不同水平ASIR-V（0、10%、30%、50%、70%、90%）算法进行图像重建。两名观察者盲法对84组CT结肠成像重组图像CT仿真内镜（CTVE）、多平面重建（MPR）、容积再现（VR）、虚拟分割（VD）]分别进行主观质量评分（4分制），同时独立测量图像噪声（SD）、信噪比（SNR）及对比噪声比（CNR）。比较图像质量主观评分一致性及不同管电流、不同水平ASIR-V重建图像的SD、SNR、CNR差异。结果 两观察者图像质量主观评分一致性好（Kappa值=0.683），管电流（r=0.734，P=0.000）及ASIR-V水平（r=0.220，P=0.044）的变化与图像质量主观评分相关，相同管电流条件下，50% ASIR-V重建图像质量主观评分最高。两观察者客观数据一致性良好。不同管电流、不同水平ASIR-V重建图像的SD、SNR及CNR差异均具有统计学意义（F=423.58、124.26、1 030.17，P<0.05）。同一管电流水平下，随着ASIR-V水平增高，图像SD降低，CNR增大；图像SNR在管电流为10、120、140、160、220、240、260 mA水平下差异具有统计学意义（F=8.75~31.36，P<0.05）。相同水平的ASIR-V重建，随管电流增大，图像SD下降，SNR及CNR逐渐升高。结论 CT结肠成像中，应用ASIR-V算法可以显著降低噪声，提高图像对比噪声比，提高图像质量，且50% ASIR-V水平在降噪能力方面更为显著。

Influence on image quality in low-dose colonography with adaptive statistical iterative reconstruction algorithm

Objective To evaluate the feasibility of adaptive statistical iterative reconstruction-V (ASIR-V)to improve image quality in low-dose CT colonography. Methods A series of thirty artificial polyps were established by ligation in an isolated segment of porcine colon. Volume data was acquired on Revolution CT scanner (GE, USA) with High Definition scan mode and different scan parameter combinations:120 kVp with different mAs (10, 30, 50, 70, 90, 100, 120, 140, 160, 180, 200, 220, 240, 260, respectively). Images were reconstructed with six different ASIR-V levels of 0(filtered back projection, FBP),10%, 30%, 50%, 70% and 90%. Two radiologists were blinded to measure and analyze the objective data independently, including image noise (SD), signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR). The kappa test was used to assess interobserver agreement in subjective image quality score. ICC test was used to examine the consistency of the measurements between two observers. SD, SNR, CNR were performed for statistical analysis in different tube current and different levels of ASIR-V using variance analysis(ANOVA). Results Interobserver agreement for subjective image quality score was good with a kappa value of 0.683. The variation of the tube current(r=0.734, P=0.000) and ASIR-V level(r=0.220, P=0.044) is related to the subjective score of image quality. Under the condition of the same tube current, image quality score of 50% ASIR-V reconstruction was the highest. Two objective data consistency is good.The differences of image noise (F=423.58, P<0.05), SNRs(F=124.26, P<0.05) and CNRs (F=1 030.17, P<0.05)of different tube current and different levels of ASIR-V reconstruction were statistically significant. In the same tube current, with increased levels of ASIR-V, image noise reduced, CNRs increased. Only in 10, 120,140, 160, 220, 240, 260 mA, the differences of SNRs were statistically significant(F=8.75-31.36, P<0.05). For the same level of ASIR-V reconstruction, with the increase of tube current, the image noise decreased, SNR and CNR increased gradually. Conclusions In the CT colonography, the application of ASIR-V algorithm can significantly reduce the noise and enhance the image contrast noise ratio and improve image quality. ASIR-V algorithm with 50% has better performance in reducing CT image noise.