Ma Xia, Zheng Bingbing, Zhu Yu, Yu Fuli, Zhang Rixin, Chen Budong
Department of Pulmonary and Critical Care Medicine, The Third Hospital of Shanxi Medical University, Taiyuan 030032, China.
Department of Pulmonary and Critical Care Medicine, The First Hospital of Shanxi Medical University, Taiyuan 030001, China.
Optik (Stuttg). 2021 Sep;241:167100. doi: 10.1016/j.ijleo.2021.167100. Epub 2021 May 7.
Since discovered in Hubei, China in December 2019, Corona Virus Disease 2019 named COVID-19 has lasted more than one year, and the number of new confirmed cases and confirmed deaths is still at a high level. COVID-19 is an infectious disease caused by SARS-CoV-2. Although RT-PCR is considered the gold standard for detection of COVID-19, CT plays an important role in the diagnosis and evaluation of the therapeutic effect of COVID-19. Diagnosis and localization of COVID-19 on CT images using deep learning can provide quantitative auxiliary information for doctors. This article proposes a novel network with multi-receptive field attention module to diagnose COVID-19 on CT images. This attention module includes three parts, a pyramid convolution module (PCM), a multi-receptive field spatial attention block (SAB), and a multi-receptive field channel attention block (CAB). The PCM can improve the diagnostic ability of the network for lesions of different sizes and shapes. The role of SAB and CAB is to focus the features extracted from the network on the lesion area to improve the ability of COVID-19 discrimination and localization. We verify the effectiveness of the proposed method on two datasets. The accuracy rate of 97.12%, specificity of 96.89%, and sensitivity of 97.21% are achieved by the proposed network on DTDB dataset provided by the Beijing Ditan Hospital Capital Medical University. Compared with other state-of-the-art attention modules, the proposed method achieves better result. As for the public COVID-19 SARS-CoV-2 dataset, 95.16% for accuracy, 95.6% for F1-score and 99.01% for AUC are obtained. The proposed network can effectively assist doctors in the diagnosis of COVID-19 CT images.
自2019年12月在中国湖北发现以来，名为COVID-19的2019冠状病毒病已持续一年多，新增确诊病例数和确诊死亡数仍处于高位。COVID-19是由严重急性呼吸综合征冠状病毒2（SARS-CoV-2）引起的一种传染病。尽管逆转录聚合酶链反应（RT-PCR）被认为是检测COVID-19的金标准，但计算机断层扫描（CT）在COVID-19的诊断和治疗效果评估中发挥着重要作用。利用深度学习对CT图像上的COVID-19进行诊断和定位可为医生提供定量辅助信息。本文提出了一种带有多感受野注意力模块的新型网络，用于在CT图像上诊断COVID-19。该注意力模块包括三个部分，即金字塔卷积模块（PCM）、多感受野空间注意力块（SAB）和多感受野通道注意力块（CAB）。PCM可以提高网络对不同大小和形状病变的诊断能力。SAB和CAB的作用是将从网络中提取的特征聚焦在病变区域，以提高COVID-19的鉴别和定位能力。我们在两个数据集上验证了所提方法的有效性。首都医科大学附属北京地坛医院提供的DTDB数据集上，所提网络实现了97.12%的准确率、96.89%的特异性和97.21%的灵敏度。与其他最先进的注意力模块相比，所提方法取得了更好的结果。对于公开的COVID-19 SARS-CoV-2数据集，准确率为95.16%，F1分数为95.6%，曲线下面积（AUC）为99.01%。所提网络可以有效地辅助医生诊断COVID-19 CT图像。
