能谱CT参数在评估局部梗死心肌图像质量中的研究

目的：通过研究评价能谱CT（Spectral CT）成像功能参数的有效性，并探讨这些参数在评估局部梗死心肌图像质量中的临床应用价值。方法：收集我院30例冠心病患者，均在一次扫描中行宝石能谱CT常规冠脉CTA（CCTA）扫描模式和宝石能谱成像（GSI）扫描模式。患者均在一周内行心脏MRI扫描。常规混合能量图像上分析冠状动脉的狭窄程度。分别找出1 min和3 min单光子能量水平的图像上梗死心肌与正常心肌的最佳对比噪声比（optimal CNR）以及对应的最佳单能量值和此能量值的图像噪声值，通过单因数方差分析方法与常规混合能量图像的对比噪声比（CNR）和图像噪声值进行比较。在MRI心肌灌注图像上找出梗死心肌并与能谱CT进行相关性对比。结果：宝石能谱CT共检出梗死心肌36个区域，所有梗死心肌在CCTA图像上均有区域供血冠脉50％以上狭窄（71.7％±16.2％），并且找出的梗死心肌与MRI心肌灌注图像有较高的相关性（r=100％，P=0.00）。1 min、3 min单光子能量水平图像上最佳CNR值明显高于常规混合能量图像CNR值（P<0.05）。1 min、3 min最佳单能量值水平的图像噪声值均低于常规混合能量图像（P<0.05）。结论：能谱CT的GSI能谱成像模式较常规混合能量图像能够明显提高图像质量，从而提高了梗死心肌的检测效率。

The Study of the Image Quality of Dual Energy Spectral CT in the Assessment of Myocardial Infarction

Objective: To study the effectiveness of the functional parameters of spectral CT, and to explore their clinic values of image quality in the evaluation of myocardial infarction (MI), so as to provide a novel and reliable imaging diagnostic system for early detection of MI. Methods: 30 patients with coronary heart disease were scanned with the conventional coronary CT angiography (CCTA) and the dual energy spectral imaging modes. In addition, all cases had the Cardiac MRI scan within a week. CCTA images were used to measure stenoses. The optimal contrast-to-noise ratio (CNR) and the optimal monochromatic energy levels for displaying infractions were measured from 1 min and 3 min monochromatic images. Record the image noise according to the optimal monochromatic energy levels. The results were compared with those measured from conventional 140kVp polychromatic images. Detect MI in MRI perfusion images and compare with those detected by CT. Results: Spectral CT have detected 36 regions of MI and had good correlation with the myocardial perfusion images of MRI (r=100%,P=0.00). All spectral CT-identified MI had associated coronary artery with more than 50% stenosis of the intravascular diameter (71.7%±16.2%). The optimal CNR for displaying infractions of 1 min and 3 min monochromatic images were significant higher than the CNR of conventional 140kVp polychromatic images (P<0.05). The image noise according to the optimal monochromatic energy levels of 1 min and 3 min monochromatic images were significant lower than conventional 140kVp polychromatic images (P<0.05). Conclusions: The GSI mode of dual energy spectral CT greatly improves the image quality and the efficiency of the detection of MI.