Myocardial Late Gadolinium Enhancement: Accuracy of T1 Mapping-based Synthetic Inversion-Recovery Imaging
Myocardial Late Gadolinium Enhancement: Accuracy of T1 Mapping-based Synthetic Inversion-Recovery Imaging
复制标题
DOI：
10.1148/radiol.2015150162
复制
发表时间：
2016-02-01
期刊：
影响因子：
19.7
通讯作者：
Schoepf, U. Joseph
中科院分区：
文献类型：
--
作者：
Varga-Szemes, Akos;van der Geest, Rob J.;Schoepf, U. Joseph
目的：比较合成反转恢复 (IR) 方法与传统 IR 技术对心肌晚期钆增强 (LGE) 的检测和定量的准确性。材料和方法：这项前瞻性研究得到了机构审查委员会的批准，并符合 HIPAA。所有患者均签署了书面知情同意书。 2014 年 6 月至 11 月期间，43 名疑似既往患有心肌梗塞的患者（25 名男性；平均年龄 54 岁+/- 16 岁）接受了磁共振 (MR) 成像，包括造影剂增强 LGE 成像和 T1 映射。合成震级和相敏红外图像是在 T1 图的基础上生成的。图像由两名读者评估。使用 McNemar 检验分析合成技术和传统技术之间每个患者和每个节段 LGE 检出率的差异，并使用配对 t 检验和 Bland-Altman 统计计算 LGE 量化的准确性。分别使用 k 和 Bland-Altman 统计数据分析读者间对 LGE 检测和量化的一致性。结果：43 名患者中有 17 名 (39%) 的 LGE 模式与心肌梗死一致。在基于患者的分析中，合成幅度和相敏 IR 技术检测 LGE 的灵敏度和特异性分别为 90% 和 95%，在基于分段的分析中分别为 94% 和 99%。使用合成红外技术测量的 LGE 面积与传统技术的面积非常一致（合成幅值和相敏红外分别为 4.35 cm(2) +/- 1.88 和 4.14 cm(2) +/- 1.62，而常规幅值和相敏 IR 分别为 4.25 cm(2) +/- 1.92 和 4.22 cm(2) +/- 1.86；P > .05）。对于 LGE 的检测（kappa > 0.81）和量化（偏差范围，-0.34 至 0.40；P > .05），读者间一致性非常好。结论：基于 T1 图谱的合成 IR 方法在既往心肌梗死患者中检测和量化心肌 LGE 的准确性与传统 IR 技术相似。使用 T1 映射来导出合成 LGE 图像可以减少成像时间和未来具有完整左心室覆盖的 T1 映射协议中对操作员的依赖。 (C) 北美放射学会，2015
Purpose: To compare the accuracy of detection and quantification of myocardial late gadolinium enhancement (LGE) with a synthetic inversion-recovery (IR) approach with that of conventional IR techniques.Materials and Methods: This prospective study was approved by the institutional review board and compliant with HIPAA. All patients gave written informed consent. Between June and November 2014, 43 patients (25 men; mean age, 54 years +/- 16) suspected of having previous myocardial infarction underwent magnetic resonance (MR) imaging, including contrast material-enhanced LGE imaging and T1 mapping. Synthetic magnitude and phase-sensitive IR images were generated on the basis of T1 maps. Images were assessed by two readers. Differences in the per-patient and per-segment LGE detection rates between the synthetic and conventional techniques were analyzed with the McNemar test, and the accuracy of LGE quantification was calculated with the paired t test and Bland-Altman statistics. Interreader agreement for the detection and quantification of LGE was analyzed with k and Bland-Altman statistics, respectively.Results: Seventeen of the 43 patients (39%) had LGE patterns consistent with myocardial infarction. The sensitivity and specificity of synthetic magnitude and phase-sensitive IR techniques in the detection of LGE were 90% and 95%, respectively, with patient-based analysis and 94% and 99%, respectively, with segment-based analysis. The area of LGE measured with synthetic IR techniques showed excellent agreement with that of conventional techniques (4.35 cm(2) +/- 1.88 and 4.14 cm(2) +/- 1.62 for synthetic magnitude and phase-sensitive IR, respectively, compared with 4.25 cm(2) +/- 1.92 and 4.22 cm(2) +/- 1.86 for conventional magnitude and phase-sensitive IR, respectively; P > .05). Interreader agreement was excellent for the detection (kappa > 0.81) and quantification (bias range, -0.34 to 0.40; P > .05) of LGE.Conclusion: The accuracy of the T1 map-based synthetic IR approach in the detection and quantification of myocardial LGE in patients with previous myocardial infarction was similar to that of conventional IR techniques. The use of T1 mapping to derive synthetic LGE images may reduce imaging times and operator dependence in future T1 mapping protocols with full left ventricular coverage. (C)RSNA, 2015