论文标题
CT扫描中肺和COVID-19病变细分的深度学习模型的全面比较
Comprehensive Comparison of Deep Learning Models for Lung and COVID-19 Lesion Segmentation in CT scans
论文作者
论文摘要
最近，在使用深度学习方法（DL）方法进行医学图像分割方面发生了爆炸。但是，由于缺乏准确/绩效评估的共同参考基础以及以前的研究使用不同的数据集进行评估，因此该领域的可靠性受到了阻碍。在本文中，提出了计算机断层扫描（CT）扫描中肺和Covid-19病变分割的DL模型的广泛比较，这也可以用作测试医学图像分割模型的基准。将四个DL体系结构（UNET，Linknet，FPN，PSPNET）与25个随机初始化和预处理编码器（VGG，Densenet，Resnet，Resnet，Resnext，resnext，resnext，dpn，MobileNet，Xception，Xception，Xeption，Xpection，Inception-V4，intection-v4，EdgitiativeNet）相结合。使用原始肺口罩进行了三个实验设置，用于肺部分割，病变分割和病变分割。一个具有100 CT扫描图像的公共COVID-19数据集（用于培训的80张，验证20个）用于培训/验证，另一个由9 CT扫描卷中的829张图像组成的不同公共数据集用于测试。提供了多个发现，包括每个实验的最佳体系结构编码器模型以及每个实验，体系结构和编码器的平均骰子结果。最后，当使用肺面膜作为预处理步骤或使用验证的模型时，上限会进行改进。提供了三个实验的源代码和600个预处理的模型，适合在没有GPU功能的实验设置中进行微调。
Recently there has been an explosion in the use of Deep Learning (DL) methods for medical image segmentation. However the field's reliability is hindered by the lack of a common base of reference for accuracy/performance evaluation and the fact that previous research uses different datasets for evaluation. In this paper, an extensive comparison of DL models for lung and COVID-19 lesion segmentation in Computerized Tomography (CT) scans is presented, which can also be used as a benchmark for testing medical image segmentation models. Four DL architectures (Unet, Linknet, FPN, PSPNet) are combined with 25 randomly initialized and pretrained encoders (variations of VGG, DenseNet, ResNet, ResNext, DPN, MobileNet, Xception, Inception-v4, EfficientNet), to construct 200 tested models. Three experimental setups are conducted for lung segmentation, lesion segmentation and lesion segmentation using the original lung masks. A public COVID-19 dataset with 100 CT scan images (80 for train, 20 for validation) is used for training/validation and a different public dataset consisting of 829 images from 9 CT scan volumes for testing. Multiple findings are provided including the best architecture-encoder models for each experiment as well as mean Dice results for each experiment, architecture and encoder independently. Finally, the upper bounds improvements when using lung masks as a preprocessing step or when using pretrained models are quantified. The source code and 600 pretrained models for the three experiments are provided, suitable for fine-tuning in experimental setups without GPU capabilities.
