OBJECTIVES:To further reduce the contrast medium (CM) dose of full aortic CT angiography (ACTA) imaging using the augmented cycle-consistent adversarial framework (Au-CycleGAN) algorithm. METHODS:We prospectively enrolled 150 consecutive patients with suspected aortic disease. All received ACTA scans of ultra-low-dose CM (ULDCM) protocol and low-dose CM (LDCM) protocol. These data were randomly assigned to the training datasets (n = 100) and the validation datasets (n = 50). The ULDCM images were reconstructed by the Au-CycleGAN algorithm. Then, the AI-based ULDCM images were compared with LDCM images in terms of image quality and diagnostic accuracy. RESULTS:The mean image quality score of each location in the AI-based ULDCM group was higher than that in the ULDCM group but a little lower than that in the LDCM group (all p < 0.05). All AI-based ULDCM images met the diagnostic requirements (score ≥ 3). Except for the image noise, the AI-based ULDCM images had higher attenuation value than the ULDCM and LDCM images as well as higher SNR and CNR in all locations of the aorta analyzed (all p < 0.05). Similar results were also seen in obese patients (BMI > 25, all p < 0.05). Using the findings of LDCM images as the reference, the AI-based ULDCM images showed good diagnostic parameters and no significant differences in any of the analyzed aortic disease diagnoses (all K-values > 0.80, p < 0.05). CONCLUSIONS:The required dose of CM for full ACTA imaging can be reduced to one-third of the CM dose of the LDCM protocol while maintaining image quality and diagnostic accuracy using the Au-CycleGAN algorithm. KEY POINTS:• The required dose of contrast medium (CM) for full ACTA imaging can be reduced to one-third of the CM dose of the low-dose contrast medium (LDCM) protocol using the Au-CycleGAN algorithm. • Except for the image noise, the AI-based ultra-low-dose contrast medium (ULDCM) images had better quantitative image quality parameters than the ULDCM and LDCM images. • No significant diagnostic differences were noted between the AI-based ULDCM and LDCM images regarding all the analyzed aortic disease diagnoses.

目的:研究全面质量管理(TQM)模型在大型医学影像设备管理中的应用,实现大型医学影像设备科学化、精细化与规范化管理.方法:针对医学影像设备、图像质量和人员管理建立TQM模型,对大型医学影像设备进行全面质量管理.选取医院临床在用的16台大型医学影像设备,按照管理方式的不同将其分为对照组和实验组,对照组采用传统管理方式,实验组采用TQM模型的管理方式,比较两组设备的专业检测情况、设备管理质量、操作人员技能评价和图像质量管理的差异性.结果:实验组设备平均故障次数低于对照组,定期全面预防性维护、记录表填写和开机率均高于对照组,差异有统计学意义(t=17.29,t=22.35,t=26.31,t=20.31;P<0.05).实验组的影像设备诊断符合率、报告合格率和图像优质率高于对照组,患者信息不一致和伪影及体位不佳情况低于对照组,差异有统计学意义(x2=423.51,x2=412.36,x2=169.74,x2=62.32,x2=322.55;P<0.05).结论:TQM模型应用于大型医学影像设备管理,可保证大型医学影像设备的稳定性、精确性和安全性,有效提高医疗服务质量,提升患者的就诊满意度.

目的:探索新型黑血动态增强核磁共振成像技术(BB DCE-MRI)在定量测量早期动脉粥样硬化炎症反应的价值.方法:20头,五指山小型猪,3～4个月龄,雌雄不限,雄性已去势,体质量(29.29±1.25)kg]行高脂高TC饲养加球囊拉伤腹主动脉以制备动脉粥样硬化模型.对模型制备前后分别进行sfQIR黑血动态增强序列扫描(TR=750 ms;TE=12 ms;TI1=325 ms;TI2=125 ms,ETL=8).扫描2层,每层15幅图像.测量模型制备前、后的临床资料(TG、TC、HDL-C、LDL-C及胰岛素水平),并计算血管功能学指标(血管壁容积传输常数(Ktrans),血浆容积(Vp),血管外细胞外容积(Ve),曲线下面积(AUC).比较(1)模型制备前后临床资料与血管形态学、功能学的相关性;(2)模型制备前后临床资料差异与血管形态学、功能学差异的相关性.相关性分析应用Spearman相关分析,差异比较应用配对资料的符号检验进行统计分析.结果:TG及TC水平与血管壁Ktrans呈明显正相关(r=0.922,P=0.000;r=0.953,P=0.000).胰岛素水平与AUC呈负相关(r=0.562,P=0.048).同时发现TG、TC的变化与Ktrans 变化有显著相关性(r=0.842,P=0.004;r=0.661,P=0.038).结论:黑血 DCE-MRI 成像方法可以反映早期动脉粥样硬化斑块炎症特征,并与临床指标相一致.可以定量测量炎性反应的影像指标为早期动脉粥样硬化提供了一种无创的影像监测手段.

目的 观察深度学习图像重建(DLIR)算法用于肥胖个体低剂量冠状动脉CT血管成像(CCTA)的可行性.方法 纳入30例体质量指数(BMI)≥28 kg/m2的疑诊冠状动脉疾病(CAD)患者,以管电压100 kV行CCTA扫描;采用DLIR(DLIR-M组及DLIR-H组)算法重建图像,并与自适应统计迭代重建(ASIR-V组)图像相比较;测量左心房(LA)、主动脉根部(Ao)、主动脉根部旁前纵隔脂肪(FAT)及各冠状动脉近端CT值、噪声及对比噪声比(CNR),以5-Likert评分比较3种重建图像质量.结果 3组图像中,AoCT值差异无统计学意义(P＞0.05).DLIR-M组、DLIR-H组图像噪声低于ASIR-V组,CNR高于ASIR-V组,其中DLIR-H组与ASIR-V组差异均有统计学意义(P均＜0.05),DLIR-M组与ASIR-V组差异均无统计学意义(P均＞0.05).DLIR-M组及DLIR-H组主观图像质量评价均明显高于ASIR-V组(P均＜0.05),DLIR-H组主观图像评分最高.30例有效辐射剂量(1.12±0.28)mSv.结论 DLIR算法可显著降低肥胖个体CCTA图像噪声,提高图像质量.

目的:探讨拉普拉斯-高斯(LOG)边缘检测算法在优化冠状动脉CT血管造影(CCTA)图像质量中的应用价值.方法:选取在医院放射科行CCTA检查的100例患者,将CCTA检查的曲面重建(CPR)图像主观检测数据纳入对照组,LOG算子边缘检测数据纳入观察组.以容积再现(VR)图像为参照标准,对比两组冠状动脉狭窄程度的诊断准确性和分段的质量评分.结果:对100例患者经VR影像诊断,确诊为轻度狭窄67例(占67％),中度狭窄24例(占24％),重度狭窄9例(9％).对照组诊断的真阳性率低于观察组,组间轻度、中度和重度狭窄真阳性率比较,差异均有统计学意义(x2=4.107,x2=5.400,x2=5.143;P<0.05);对照组右冠状动脉、左回旋支动脉、左前降支和冠状动脉的质量评分低于观察组,组间差异均有统计学意义(t=2.126,t=2.333,t=2.183,t=2.762;P<0.05).结论:LOG边缘检测算法可提高冠状动脉CCTA图像的清晰度和血管边缘锐利度,血管边缘信息连贯且对比度高,其在优化CCTA图像质量中具有较高的临床应用价值.

Purpose: The objective of this study was to investigate whether three dimentional (3D)Coronary CT angiography (CCTA)feature tracking (FT) can measure global myocardial strain of the left ventricle (LV) in patients with heart failure using cardiac MR (CMR) as reference. Methods: Consecutive patients (n = 44) with variable degrees of heart failure who underwent an ECG-gated CCTA and CMR within 24 h were included. Both modalities were compared for 2D/3D LV global radial strain (2D/3DGRS), circumferential strain (2D/3D-GCS), longitudinal strain (2D/3D-GLS) and conventional functional parameters. Results: Compared to CMR, CCTA-derived 3D-GLS and LVEF showed no significant difference (p > 0.05). Bland Altman plots showed a small bias (0.3 %) between CCTA-derived 3D-GLS and CMR 3D-GLS. Close correlations were observed between the two modalities regarding LV global strain (3D-GRS, r = 0.89; 3D-GCS, r = 0.86; 3DGLS, r = 0.79, respectively, p < 0.001 for all). However, CCTA-derived 3D-GRS and 3D-GCS were statistically different compared with CMR. CCTA-derived 3D-GLS had an inverse correlation with CCTA-LVEF(r=-0.75, p < 0.05). Intraobserver agreements for CCTA-derived 3D-global strain were good (ICC = 0.856 for 3D-GLS, ICC = 0.741 for 3D-GCS and ICC = 0.762 for 3D-GRS). 2D global strain showed statistical differences between the two modalities (p<0.05 for all), but close correlations were observed regarding 2D LV global strain (2D-GRS, r = 0.80; 2D-GCS, r = 0.81; 2D-GLS, r = 0.81, respectively, p < 0.001 for all). The average radiation dose-long product (DLP) of CCTA was 387.86 +/- 89.3 mGy*cm. Conclusion: CCTA-derived 3D-GLS can provide both reliable and interchangeable results for quantitative assessment of myocardial mechanical changes in HF patients compared to CMR with good intra-observer agreement.