【摘要】 目的 分析西门子双源双能量CT(Dual-Energy CT,DECT)不同卷积函数(kernel)在痛风检测中对足踝部尿酸单钠晶体(Monosodium Urate Crystal,MSU)的显示效果。方法 分别在正弦图确定迭代算法(Sinogram-Affirmed Iterative Reconstruction,SAFIRE)、滤波反投影法(Filtered-Back Projection,FBP)下使用不同kernels(SAFIRE3:Q30、Q34、Q40;FBP:D24f、D30f、D34f、D36f、D40f)重建22例临床诊断为痛风的患者足踝部DECT影像,使用Gout软件行痛风检测。以超声结果作为参考标准,使用方差分析和秩和检验比较不同kernels对MSU体积、MSU数量、伪影数量、图像背景噪声、信噪比(Signal to Noise Ratio,SNR)的影响,并分析测量参数与患者血尿酸(Serum Uric Acid,SUA)水平、病程的相关性。结果 22例患者超声均可见MSU沉积,除SAFIRE3 Q34、FBP D24f、FBP D34f外,其余各kernels重建图像均存在不同程度的漏诊。所有图像均存在MSU伪影,以韧带/肌腱、甲床、肌肉内多见,SAFIRE3 Q34的MSU伪影总数少于FBP D24f、FBP D34f(P<0.05),其肌肉内和韧带/肌腱伪影数量与FBP D24f比较,差异无统计学意义(P>0.05),但少于FBP D34f(P<0.05)。不同kernels图像测得的MSU总体积、SNR存在差异(P<0.05),MSU数量、图像背景噪声比较,差异均无统计学意义(P>0.05)。SAFIRE3 Q34图像SNR与FBP D24f比较,差异无统计学意义(P>0.05),但优于FBP D34f(P<0.05)。伪影数量、MSU体积、MSU数量与血SUA水平、病程均无相关性。结论 使用SAFIRE3 Q34重建图像用于足踝部DECT痛风检测,可获得较好的图像质量、减少伪影,并避免漏诊。更多还原

【Abstract】 Objective To analyze the influence of different convolution functions(kernel) on the display of monosodium urate crystal(MSU) of ankle joint in gout detection by Siemens dual-energy CT(DECT). Methods Different kernels(SAFIRE3: Q30, Q34,Q40; FBP: D24f, D30f, D34f, D36f, D40f) reconstructed the ankle joint DECT images of 22 clinically diagnosed patients with gout,and was detected by Gout software. Using ultrasound results as reference standard, the effects of different kernels on MSU volume,MSU number, artifact number, image background noise and signal to noise ratio(SNR) were compared by variance analysis and rank-sum test, and the correlation between measurement parameters and serum uric acid(SUA) level and course of disease was analyzed. Results MSU deposition was found in all the 22 patients by ultrasound. Except SAFIRE3 Q34, FBP D24f and FBP D34f,the other kernels reconstructed images had different degrees of missed diagnosis. MSU artifacts were found in all the reconstructed images, mostly distributed in ligaments/tendons, nail beds and muscles. The total number of MSU artifacts in SAFIRE3 Q34was less than that in FBP D24f and FBP D34f(P<0.05). The number of intramuscular and ligament/tendon artifacts was similar to that of FBP D24f(P>0.05), but less than that of FBP D34f(P<0.05). There were significant differences in MSU volume and SNR between kernels images(P<0.05), while there were no significant differences in MSU number and image back graud noise(P>0.05). The SNR of SAFIRE3 Q34 image was similar to FBP D24f(P>0.05) and superior to FBP D34f(P<0.05). There was no correlation between the number of artifacts, the MSU volume and MSU number and the level of SUA or the course of disease.Conclusion SAFIRE3 Q34 reconstruction image was recommended for DECT gout detection of ankle joint to obtain better image quality, reduce artifacts, and avoid missed diagnosis.更多还原