可形变Transformer辅助的胸部X光影像疾病诊断模型  

Chest X-ray imaging disease diagnosis model assisted bydeformable Transformer

作　　者：胡锦波 聂为之 宋丹 高卓 白云鹏 赵丰[3] HU Jin-bo;NIE Wei-zhi;SONG Dan;GAO Zhuo;BAI Yun-peng;ZHAO Feng(School of Electrical and Information Engineering,Tianjin University,Tianjin 300072,China;School of Information,Changchun Polytechnic,Changchun 130033,China;Department of Cardiovascular Surgery,Tianjin Chest Hospital,Tianjin 300222,China)

机构地区：[1]天津大学电气自动化与信息工程学院,天津300072 [2]长春职业技术学院信息学院,吉林长春130033 [3]天津市胸科医院心血管外科,天津300222

出　　处：《浙江大学学报（工学版）》2023年第10期1923-1932,共10页Journal of Zhejiang University：Engineering Science

基　　金：国家自然科学基金资助项目(61902277,62272337).

摘　　要：针对胸部X光影像中的灰雾现象、病变区域重叠等问题,提出可形变Transformer辅助的胸部X光影像疾病诊断模型.将扩展后的ResNet50作为特征提取网络,添加压缩型双注意力模块,增强病变区域与非病变区域之间的特征差异,降低冗余信息的干扰,提高图像数据的特征提取效果;通过可形变Transformer解码器内部的交叉注意力模块,引入类别表征作为先验知识,引导影像特征进一步融合,提高不同疾病在影像区域重叠情况下的特征区分度;将解码器的输出传入分类器中以获得最终的诊断结果.压缩型双注意力模块和可形变Transformer均起到降低模型计算复杂度的作用,引入非对称损失函数可以更好地解决正负样本不均衡.利用所提模型在公开数据集ChestX-Ray14和CheXpert上进行多组实验,在2个数据集上的受试者操作的特征曲线下面积值(AUC)分别达到0.8398和0.9061,表明该模型在胸部X光影像的疾病诊断方面具有正确性和有效性.A disease diagnosis model for chest X-ray images assisted by deformable Transformer was proposed,aiming at the problems of gray fog phenomenon and overlapping lesion areas in chest X-ray images.The extended residual network ResNet50 was used as a feature extraction network.A compressed dual attention module was added to enhance the feature difference between the lesion area and the non-lesion area,further reduced the interference of redundant information and improved the feature extraction of image data.Through the cross-attention module inside the deformable Transformer decoder,category representations were introduced as the priori knowledge to guide further fusion of image features and improve the feature discrimination of different diseases in the case of overlapping image regions.Output of the decoder was passed into the classifier to obtain the final diagnosis.Both the compressed dual attention module and the deformable Transformer can reduce the computational complexity of the model.The asymmetric loss function was introduced to solve the imbalance of positive and negative samples.The proposed model was subjected to multiple sets of experiments on public datasets ChestX-Ray14 and CheXpert.The area under curve(AUC)on two datasets reached 0.8398 and 0.9061 respectively,indicating the correctness and validity of the model for disease diagnosis on chest X-ray images.

关 键 词：胸部X光图像分类 可形变Transformer 压缩型双注意力 非对称损失函数 先验知识 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]