QGS和ECTb软件在门控心肌灌注显像定量分析中的比较

Comparison of quantitative analysis between QGS and ECTb software programs used in gated myocardial perfusion imaging

目的·探讨单光子发射计算机断层成像(single-photon emission computed tomography,SPECT)门控心肌灌注显像(gated myocardial perfusion imaging,G-MPI)中常用的2种处理软件Cedars-Sinai定量门控SPECT(quantitative gated SPECT,QGS)和爱莫瑞心脏工具箱(Emory cardiac toolbox,ECTb)所得定量分析结果的差异。方法·回顾性分析以锝-99m-甲氧基异丁基异腈(99mTcmethoxyisobutylisonitrile,99mTc-MIBI)为心肌灌注显像剂进行SPECT G-MPI的28例患者(共37人次)的检查结果,使用QGS和ECTb 2种后处理软件分别对心功能参数[左心室射血分数(left ventricular ejection fraction,LVEF)、舒张末期容积(end-diastolic volume,EDV)和收缩末期容积(end-systolic volume,ESV)]及相位分析参数[相位直方图带宽(phase histogram bandwidth,PHB)和相位标准差(phase standard deviation,PSD)]进行定量分析。并分析同一参数在2种软件处理结果中的相关性及差异。结果·2种软件所测得LVEF、EDV和ESV值的相关性较好(LVEF:r=0.917,P=0.000。EDV:r=0.976,P=0.000。ESV:r=0.981,P=0.000)。但2种软件进行相位分析所得PHB及PSD的相关性均不明显(PHB:r=0.319,P=0.055。PSD:r=0.172,P=0.310)。在心功能分析方面,QGS所测得ESV值高于ECTb所测得值,而QGS所测得EDV、LVEF值则均低于ECTb所测得值。在相位分析方面,QGS所测得PSD、PHB值均低于ECTb所测得值。分析2种软件所测结果之间的差异,发现2种软件所测得LVEF、ESV、PSD值的差异有统计学意义[LVEF:(47.8±16.9)%vs (57.4±17.2)%,P=0.000。ESV:(67.5±51.0) m L vs (58.3±50.0) mL,P=0.000。PSD:20.5o±10.3ovs 30.6o±18.9o, P=0.004]。而2种软件所测得EDV、PHB值的差异无统计学意义[EDV:(116.8±52.8) m L vs(120.8±55.7) mL,P=0.050。PHB:72.2o±37.0ovs 86.1o±55.7o,P=0.139]。结论·QGS与ECTb软件在G-MPI定量分析中所测得左心功能参数的相关性好,但2种软件在左心室机械不同步评估(相位分析)中无明显相关性。2种软件所测得同一参数的值之间有差异,故在临床应用中2种软件所测得结果不适合直接对比参考。2种软件所测得结果的差异提示,在临床工作中有必要根据科室所用软件建立本单位正常人群的数据库。

Objective · To study the differences and correlations of quantitative analysis between Cedars-Sinai quantitative gated SPECT （QGS） and Emory cardiac toolbox （ECTb） used in single photon emission computed tomography （SPECT） gated myocardial perfusion imaging （G-MPI）. Methods · A total of 28 patients were examined with 99mTc-methoxyisobutylisonitrile （99mTc-MIBI） SPECT G-MPI. The left ventricular ejection fraction （LVEF）, end-diastolic volume （EDV）, end-systolic volume （ESV）, phase histogram bandwidth （PHB） and phase standard deviation （PSD） were calculated with QGS and ECTb. The correlations and differences of the results from these two programs were analyzed. Results · These two software programs showed high correlation for LVEF, EDV and ESV （LVEF r=0.917, P=0.000. EDV： r=0.976, P=0.000. ESV： r=0.981, P=0.000）. The analysis showed no signifcant correlation for PHB and PSD （PHB： r=0.319, P=0.055. PSD： r=0.172, P=0.310）. In the analysis of cardiac function, the ESV measured by QGS was higher than that measured by ECTb, and the EDV and LVEF were lower than those measured by ECTb. In the phase analysis, the PSD and PHB measured by QGS were lower than those measured by ECTb. These differences between the results measured by the two software programs were not consistency. There were signifcant differences in LVEF, ESV and PSD in the comparison of QGS and ECTb [LVEF： （47.8±16.9）% vs （57.4±17.2）%, P=0.000. ESV： （67.5±51.0） mL vs （58.3±50.0） mL, P=0.000. PSD： 20.5o±10.3o vs 30.6o±18.9o, P =0.004]. The EDV and PHB showed no signifcant difference between the QGS and ECTb [EDV： （116.8±52.8） mL vs （120.8±55.7） mL, P=0.050. PHB： 72.2o±37.0o vs 86.1o±55.7o, P=0.139]. Conclusion · These two software programs have good consistency in quantitative analysis of cardiac function. But the result shows no significant consistent in the evaluation of left ventricular mechanical dyssynchrony. There are differences between the data measured by QGS and ECTb. Using the results measured by the two software programs for direct comparison may be not suitable in clinical applications. The differences between these two software programs indicate that it may be necessary to establish a normal databases in clinical work based on the local conditions.