背景:开发了一种称为MUS方法(未平滑图像上的掩蔽过程)的新图像重建过程,以消除伪影,尤其是那些在下墙的人。我们比较了MUS和常规方法在应激心肌灌注SPECT(MPS)中的诊断性能。
方法:纳入126例患者,这些患者接受了99mTc-MIBI的压力休息MPS。患者分为两组:91例RCA或LCX狭窄<50%(非缺血组)和35例RCA狭窄≥90%或FFR阳性(缺血组),根据MPS3个月内的冠状动脉CT或冠状动脉造影检查。当对应于下壁区域的五个节段的总和差异评分≥2时,缺血性心脏病(IHD)被认为是阳性。
结果:MUS方法与常规方法的灵敏度相当(有序子集期望最大化;OSEM)(51%对54%,分别;(p=0.366),使用MUS方法的特异性显著更高(87%vs77%,分别为;p<0.05),使用MUS方法的诊断性能更高(曲线下面积[AUC],常规0.61vs.MUS0.69,p=0.138)。在排除39名接受额外俯卧成像的患者后,对87名患者进行了评估,使用MUS方法的灵敏度为44%,与常规方法的44%相当,但特异性为90%,显著高于常规方法的77%(p<0.05)。MUS方法的诊断性能较高(AUC,常规0.60vs.MUS0.67,p=0.185)。
结论:使用MUS方法提高了诊断IHD的特异性,同时保持了敏感性,与常规方法相比。MUS方法可以实现与仰卧位相当的诊断准确性的提高,特别是在执行俯卧位困难的患者中,不增加患者负担。
BACKGROUND: A new image reconstruction process termed the
MUS method (masking process on unsmoothed images) was developed to eliminate artifacts, especially those in the inferior wall. We compared diagnostic performance between the
MUS and conventional method in stress myocardial perfusion SPECT (MPS).
METHODS: Enrolled were 126 patients who underwent stress-rest MPS with 99 m Tc-MIBI. Patients were divided into two groups: 91 with < 50% stenosis in the RCA or LCX (non-ischemia group) and 35 patients with ≥ 90% stenosis or FFR-positive in the RCA (ischemia group), according to coronary CT or coronary angiography within 3 months of MPS. Ischemic heart disease (IHD) was considered positive when the summed difference score of five segments corresponding to the inferior wall region was ≥ 2.
RESULTS: Sensitivity was comparable between the MUS method and the conventional method (ordered subset expectation maximization; OSEM) (51% vs 54%, respectively; (p = 0.366), specificity was significantly higher using the MUS method (87% vs 77%, respectively; p < 0.05), and diagnostic performance was higher using the MUS method (area under curve [AUC], conventional 0.61 vs.
MUS 0.69, p = 0.138). In evaluation of 87 patients after excluding 39 who received additional prone imaging, sensitivity using the MUS method was 44%, which was comparable to 44% using the conventional method but specificity was 90%, which was significantly higher than 77% using the conventional method (p < 0.05). The diagnostic performance of the
MUS method was higher (AUC, conventional 0.60 vs.
MUS 0.67, p = 0.185).
CONCLUSIONS: Use of the MUS method improved specificity in diagnosis of IHD while maintaining sensitivity, compared with the conventional method. The MUS method can achieve an improvement in diagnostic accuracy equivalent to the supine position, particularly in patients who have difficulty performing the prone position, without increasing the patient burden.