Filipczak Krzysztof, Kuśmierek Jacek, Drożdż Jarosław, Chiżyński Krzysztof, Kasprzak Jarosław D, Peruga Jan Z, Płachcińska Anna
Department of Quality Control and Radiological Protection, Inter-Faculty Chair of Medical Imaging and Radiology, Medical University of Lodz, Poland.
Nucl Med Rev Cent East Eur. 2017;20(1):25-31. doi: 10.5603/NMR.a2016.0041.
The work compares the results of shortened gated myocardial perfusion imaging (MPI), processed with "Myovation Evolution" software, with a study performed in a standard way.
A retrospective study was conducted in a group of 95 patients (56 males and 39 females, age 62 ± 9 years, BMI 28 ± 4) with known or suspected CAD, without clinical history or any signs of a previous myocardial infarction. All patients underwent coronary angiography (CA) within 3 months of MPI. CA was used as a reference for diagnostic performance of MPI. Patients underwent a stress/rest 2-day MPI. Both studies were performed twice, with normal (25s) and shortened (13s) time per projection. Studies were processed using Myovation protocol (OSEM with 2 iterations and 10 subsets) for full time (FT) studies and a Myovation Evolution protocol dedicated to half time (HT) studies (OSEM with 12 iterations, 10 subsets). Reconstructed images, with and without attenuation correction (AC), were evaluated by 2 experienced nuclear medicine specialists (a consensus), with regard to image quality and perfusion, evaluated using a visual semi-quantitative method, applying a standard division of myocardium into 17 segments. Perfusion was assessed in every segment using a standard 5 grade scale. Summed stress scores were calculated for every patient and threshold values for detection of CAD were selected based on ROC analysis with CA treated as a reference method. After 2 months FT images were interpreted again by the same specialists.
The quality of images obtained from shortened and normal studies was equally good. All correlation coefficients between segmental scorings of FT and HT studies were high and statistically significant. Correlation coefficients between corresponding segments in FTAC and HTAC (i.e. with AC) studies were systematically higher than without AC. The agreement between FT and HT study results was equal to 81% for FT and HT studies and to 86% for FTAC and HTAC studies (p = 0.40). The repeatability of FTAC study assessments was equal to 94%. 95-percent confidence intervals calculated for agreement between FTAC and HTAC studies and the repeatability of FTAC study overlapped considerably. Correlation coefficients for EDV, ESV and EF values between FT and HT were high: 0.93, 0.96 and 0.88, respectively.
Myovation Evolution protocol used for reconstruction of myocardial perfusion studies with reduced number of counts requires AC. The agreement between the results of visual assessment of normal and reduced count studies is high and not worse than the agreement between repeat assessment of a full time study.
本研究比较了采用“Myovation Evolution”软件处理的缩短门控心肌灌注成像（MPI）结果与采用标准方法进行的研究结果。
对95例已知或疑似冠心病患者（56例男性，39例女性，年龄62±9岁，BMI 28±4）进行回顾性研究，这些患者无临床病史或既往心肌梗死迹象。所有患者在MPI检查后3个月内接受冠状动脉造影（CA）。CA用作MPI诊断性能的参考。患者接受了负荷/静息2日MPI检查。两项研究均进行两次，每次投影时间分别为正常（25秒）和缩短（13秒）。采用Myovation协议（2次迭代和10个子集的有序子集期望最大化算法）处理全时（FT）研究，采用专门用于半时（HT）研究的Myovation Evolution协议（12次迭代，10个子集的有序子集期望最大化算法）。由2名经验丰富的核医学专家（达成共识）对有无衰减校正（AC）的重建图像进行图像质量和灌注评估，采用视觉半定量方法，将心肌标准划分为17个节段。使用标准的5级评分量表评估每个节段的灌注情况。计算每位患者的负荷总分，并以CA作为参考方法，通过ROC分析选择CAD检测的阈值。2个月后，同一位专家再次解读FT图像。
缩短研究和正常研究获得的图像质量同样良好。FT和HT研究节段评分之间的所有相关系数均较高且具有统计学意义。FTAC和HTAC（即有AC）研究中相应节段之间的相关系数系统地高于无AC的情况。FT和HT研究结果之间的一致性在FT和HT研究中为81%，在FTAC和HTAC研究中为86%（p = 0.40）。FTAC研究评估的可重复性为94%。FTAC和HTAC研究之间一致性以及FTAC研究可重复性计算的95%置信区间有相当大的重叠。FT和HT之间EDV、ESV和EF值的相关系数较高：分别为0.93、0.96和0.88。
用于重建计数减少的心肌灌注研究的Myovation Evolution协议需要AC。正常计数研究和减少计数研究的视觉评估结果之间的一致性较高，且不低于全时研究重复评估之间的一致性。
