Improving the Generalizability of Convolutional Neural Network-Based Segmentation on CMR Images
Improving the Generalizability of Convolutional Neural Network-Based Segmentation on CMR Images
复制标题
DOI：
10.3389/fcvm.2020.00105
复制
发表时间：
2020-06-30
影响因子：
3.6
通讯作者：
Rueckert, Daniel
中科院分区：
文献类型：
--
作者：
Chen, Chen;Bai, Wenjia;Rueckert, Daniel
背景：基于卷积神经网络（CNN）的分割方法为临床医生评估心脏 MR 图像中心脏的结构和功能提供了一种有效且自动化的方法。虽然当训练和测试图像来自同一域（例如，相同的扫描仪或站点）时，CNN 通常可以高精度地执行分割任务，但它们的性能通常会在来自不同扫描仪或临床站点的图像上显着下降。方法：我们提出了一种简单而有效的方法，通过仔细设计数据标准化和增强策略来适应多站点、多扫描仪临床成像数据集中的常见场景，以提高网络泛化能力。我们证明，在英国生物银行的单站点单扫描仪数据集上训练的神经网络可以成功应用于跨不同站点和不同扫描仪分割心脏 MR 图像，而不会严重损失准确性。具体来说，该方法在英国生物银行的 3,975 名受试者中进行了训练。然后直接在英国生物银行的 600 名不同受试者上进行域内测试，并在另外两个数据集上进行跨域测试：ACDC 数据集（100 名受试者、1 个站点、2 个扫描仪）和 BSCMR-AS 数据集（599 名受试者、6 个站点、9 个扫描仪）。 结果：所提出的方法在英国生物库测试集上产生了有希望的分割结果，与文献中先前报告的值相当，同时在跨域测试集上也表现良好，达到了平均水平ACDC 数据集上左心室的 Dice 指标为 0.90，心肌的 Dice 指标为 0.81，右心室的 Dice 指标为 0.82；在 BSCMR-AS 数据集上，左心室为 0.89，心肌为 0.83。 结论：所提出的方法为提高跨扫描仪和跨站点心脏 MR 图像分割任务的基于 CNN 的模型泛化性提供了潜在的解决方案。
Background: Convolutional neural network (CNN) based segmentation methods provide an efficient and automated way for clinicians to assess the structure and function of the heart in cardiac MR images. While CNNs can generally perform the segmentation tasks with high accuracy when training and test images come from the same domain (e.g., same scanner or site), their performance often degrades dramatically on images from different scanners or clinical sites.Methods: We propose a simple yet effective way for improving the network generalization ability by carefully designing data normalization and augmentation strategies to accommodate common scenarios in multi-site, multi-scanner clinical imaging data sets. We demonstrate that a neural network trained on a single-site single-scanner dataset from the UK Biobank can be successfully applied to segmenting cardiac MR images across different sites and different scanners without substantial loss of accuracy. Specifically, the method was trained on a large set of 3,975 subjects from the UK Biobank. It was then directly tested on 600 different subjects fromthe UK Biobank for intra-domain testing and two other sets for cross-domain testing: the ACDC dataset (100 subjects, 1 site, 2 scanners) and the BSCMR-AS dataset (599 subjects, 6 sites, 9 scanners).Results: The proposed method produces promising segmentation results on the UK Biobank test set which are comparable to previously reported values in the literature, while also performing well on cross-domain test sets, achieving a mean Dice metric of 0.90 for the left ventricle, 0.81 for the myocardium, and 0.82 for the right ventricle on the ACDC dataset; and 0.89 for the left ventricle, 0.83 for the myocardium on the BSCMR-AS dataset.Conclusions: The proposed method offers a potential solution to improve CNN-based model generalizability for the cross-scanner and cross-site cardiac MR image segmentation task.