Stanford Medicine, Pasteur Drive 300, Stanford, CA, 94305, USA.
National Heart Lung and Blood Institute, National Institutes of Health, Bethesda, MD, USA.
Eur Radiol. 2021 Jun;31(6):3941-3950. doi: 10.1007/s00330-020-07474-5. Epub 2020 Nov 27.
Cardiac magnetic resonance (CMR) first-pass perfusion is an established noninvasive diagnostic imaging modality for detecting myocardial ischemia. A CMR perfusion sequence provides a time series of 2D images for dynamic contrast enhancement of the heart. Accurate myocardial segmentation of the perfusion images is essential for quantitative analysis and it can facilitate automated pixel-wise myocardial perfusion quantification.
In this study, we compared different deep learning methodologies for CMR perfusion image segmentation. We evaluated the performance of several image segmentation methods using convolutional neural networks, such as the U-Net in 2D and 3D (2D plus time) implementations, with and without additional motion correction image processing step. We also present a modified U-Net architecture with a novel type of temporal pooling layer which results in improved performance.
The best DICE scores were 0.86 and 0.90 for LV myocardium and LV cavity, while the best Hausdorff distances were 2.3 and 2.1 pixels for LV myocardium and LV cavity using 5-fold cross-validation. The methods were corroborated in a second independent test set of 20 patients with similar performance (best DICE scores 0.84 for LV myocardium).
Our results showed that the LV myocardial segmentation of CMR perfusion images is best performed using a combination of motion correction and 3D convolutional networks which significantly outperformed all tested 2D approaches. Reliable frame-by-frame segmentation will facilitate new and improved quantification methods for CMR perfusion imaging.
• Reliable segmentation of the myocardium offers the potential to perform pixel level perfusion assessment. • A deep learning approach in combination with motion correction, 3D (2D + time) methods, and a deep temporal connection module produced reliable segmentation results.
心脏磁共振（CMR）首过灌注是一种成熟的非侵入性诊断成像方式，用于检测心肌缺血。CMR 灌注序列提供一系列 2D 图像，用于心脏的动态对比增强。准确的灌注图像心肌分段对于定量分析至关重要，它可以促进自动像素级心肌灌注定量。
在这项研究中，我们比较了用于 CMR 灌注图像分割的不同深度学习方法。我们使用卷积神经网络评估了几种图像分割方法的性能，例如 2D 和 3D（2D 加时间）实现的 U-Net，以及是否具有附加的运动校正图像处理步骤。我们还提出了一种具有新型时间池化层的改进 U-Net 架构，从而提高了性能。
使用 5 折交叉验证，LV 心肌和 LV 腔的最佳 DICE 评分分别为 0.86 和 0.90，而 LV 心肌和 LV 腔的最佳 Hausdorff 距离分别为 2.3 和 2.1 像素。在第二个独立的 20 例患者测试集中，这些方法得到了类似的验证，LV 心肌的最佳 DICE 评分为 0.84。
我们的结果表明，CMR 灌注图像的 LV 心肌分割最好使用运动校正和 3D 卷积网络的组合来完成，这明显优于所有测试的 2D 方法。可靠的逐帧分割将为 CMR 灌注成像的新的和改进的定量方法提供便利。
• 心肌的可靠分割有可能进行像素级灌注评估。
• 结合运动校正、3D（2D+时间）方法和深度时间连接模块的深度学习方法产生了可靠的分割结果。
