Thorax disease classification with attention guided convolutional neural network
Thorax disease classification with attention guided convolutional neural network
复制标题
DOI：
10.1016/j.patrec.2019.11.040
复制
发表时间：
2020-03-01
影响因子：
5.1
通讯作者：
Yang, Yi
中科院分区：
文献类型：
--
作者：
Guan, Qingji;Huang, Yaping;Yang, Yi
本文研究了胸部X线（CXR）图像的胸部疾病诊断任务。大多数现有方法通常以全局图像作为输入来学习网络。然而，胸部疾病通常发生在疾病特异性的（小）局部区域。因此，使用全局图像训练CNN可能会受到（过多）不相关噪声区域的影响。此外，由于一些CXR图像对齐不良，不规则边界的存在阻碍了网络性能。为了解决上述问题，我们提出将全局和局部线索集成到三分支注意力引导卷积神经网络（AG-CNN）中以识别胸部疾病。提出了一种基于注意引导掩码推理的裁剪策略，以避免噪声和提高全局分支的对齐。AG-CNN还整合了全局线索，以补偿局部分支丢失的判别线索。具体来说，我们首先使用全局图像学习全局CNN分支。然后，在从全局分支生成的注意力热图的引导下，我们推断出掩模以从全局图像中裁剪出有区别的区域。局部区域用于训练局部CNN分支。最后，我们将全局分支和局部分支的最后一个池化层连接起来，以微调融合分支。在ChestX-ray 14数据集上的实验表明，在将局部线索与全局信息整合后，AG-CNN提高了平均AUC分数。(C)2019爱思唯尔B. V.保留所有权利。
This paper considers the task of thorax disease diagnosis on chest X-ray (CXR) images. Most existing methods generally learn a network with global images as input. However, thorax diseases usually happen in (small) localized areas which are disease specific. Thus training CNNs using global images may be affected by the (excessive) irrelevant noisy areas. Besides, due to the poor alignment of some CXR images, the existence of irregular borders hinders the network performance. For addressing the above problems, we propose to integrate the global and local cues into a three-branch attention guided convolution neural network (AG-CNN) to identify thorax diseases. An attention guided mask inference based cropping strategy is proposed to avoid noise and improve alignment in the global branch. AG-CNN also integrates the global cues to compensate the lost discriminative cues by the local branch. Specifically, we first learn a global CNN branch using global images. Then, guided by the attention heatmap generated from the global branch, we infer a mask to crop a discriminative region from the global image. The local region is used for training a local CNN branch. Lastly, we concatenate the last pooling layers of both the global and local branches for fine-tuning the fusion branch. Experiments on the ChestX-ray14 dataset demonstrate that after integrating the local cues with the global information, the average AUC scores are improved by AG-CNN. (C) 2019 Elsevier B.V. All rights reserved.