Kero Tanja, Nordström Jonny, Harms Hendrik J, Sörensen Jens, Ahlström Håkan, Lubberink Mark
Nuclear Medicine & PET, Uppsala University, Uppsala, Sweden.
PET Center/Medical Imaging Center, Uppsala University Hospital, 75185, Uppsala, Sweden.
EJNMMI Phys. 2017 Dec;4(1):1. doi: 10.1186/s40658-016-0171-2. Epub 2017 Jan 6.
The use of integrated PET-MR offers new opportunities for comprehensive assessment of cardiac morphology and function. However, little is known on the quantitative accuracy of cardiac PET imaging with integrated time-of-flight PET-MR. The aim of the present work was to validate the GE Signa PET-MR scanner for quantitative cardiac PET perfusion imaging. Eleven patients (nine male; mean age 59 years; range 46-74 years) with known or suspected coronary artery disease underwent O-water PET scans at rest and during adenosine-induced hyperaemia on a GE Discovery ST PET-CT and a GE Signa PET-MR scanner. PET-MR images were reconstructed using settings recommended by the manufacturer, including time-of-flight (TOF). Data were analysed semi-automatically using Cardiac VUer software, resulting in both parametric myocardial blood flow (MBF) images and segment-based MBF values. Correlation and agreement between PET-CT-based and PET-MR-based MBF values for all three coronary artery territories were assessed using regression analysis and intra-class correlation coefficients (ICC). In addition to the cardiac PET-MR reconstruction protocol as recommended by the manufacturer, comparisons were made using a PET-CT resolution-matched reconstruction protocol both without and with TOF to assess the effect of time-of-flight and reconstruction parameters on quantitative MBF values.
Stress MBF data from one patient was excluded due to movement during the PET-CT scanning. Mean MBF values at rest and stress were (0.92 ± 0.12) and (2.74 ± 1.37) mL/g/min for PET-CT and (0.90 ± 0.23) and (2.65 ± 1.15) mL/g/min for PET-MR (p = 0.33 and p = 0.74). ICC between PET-CT-based and PET-MR-based regional MBF was 0.98. Image quality was improved with PET-MR as compared to PET-CT. ICC between PET-MR-based regional MBF with and without TOF and using different filter and reconstruction settings was 1.00.
PET-MR-based MBF values correlated well with PET-CT-based MBF values and the parametric PET-MR images were excellent. TOF and reconstruction settings had little impact on MBF values.
一体化PET-MR的应用为全面评估心脏形态和功能提供了新的机遇。然而,关于一体化飞行时间PET-MR心脏PET成像的定量准确性,人们了解甚少。本研究的目的是验证GE Signa PET-MR扫描仪用于心脏PET定量灌注成像的性能。11例已知或疑似冠心病患者(9例男性;平均年龄59岁;范围46-74岁)在GE Discovery ST PET-CT和GE Signa PET-MR扫描仪上进行了静息和腺苷诱发充血状态下的O水PET扫描。PET-MR图像使用制造商推荐的设置进行重建,包括飞行时间(TOF)。使用Cardiac VUer软件进行半自动数据分析,生成参数化心肌血流量(MBF)图像和基于节段的MBF值。使用回归分析和组内相关系数(ICC)评估基于PET-CT和基于PET-MR的MBF值在所有三个冠状动脉区域之间的相关性和一致性。除了制造商推荐的心脏PET-MR重建方案外,还使用PET-CT分辨率匹配的重建方案进行了比较,该方案分别在不使用和使用TOF的情况下进行,以评估飞行时间和重建参数对定量MBF值的影响。
由于PET-CT扫描期间的运动,排除了一名患者的应激MBF数据。PET-CT静息和应激状态下的平均MBF值分别为(0.92±0.12)和(2.74±1.37)mL/g/min,PET-MR分别为(0.90±0.23)和(2.65±1.15)mL/g/min(p=0.33和p=0.74)。基于PET-CT和基于PET-MR的区域MBF之间的ICC为0.98。与PET-CT相比,PET-MR的图像质量有所提高。基于PET-MR的区域MBF在使用和不使用TOF以及使用不同滤波和重建设置之间的ICC为1.00。
基于PET-MR的MBF值与基于PET-CT的MBF值相关性良好,参数化PET-MR图像质量优异。TOF和重建设置对MBF值影响较小。