关键词:
computer-aided diagnosis(CAD)
medical image classification
deep learning
feature symmetry
mirror loss
摘要:
Computer-aided diagnosis(CAD)can detect tuberculosis(TB)cases,providing radiologists with more accurate and efficient diagnostic *** noise information in TB chest X-ray(CXR)images is a major challenge in this classification *** study aims to propose a model with high performance in TB CXR image detection named multi-scale input mirror network(MIM-Net)based on CXR image symmetry,which consists of a multi-scale input feature extraction network and mirror *** multi-scale image input can enhance feature extraction,while the mirror loss can improve the network performance through *** used a publicly available TB CXR image classification dataset to evaluate our proposed method via 5-fold cross-validation,with accuracy,sensitivity,specificity,positive predictive value,negative predictive value,and area under curve(AUC)of 99.67%,100%,99.60%,99.80%,100%,and 0.9999,*** to other models,MIM-Net performed best in all ***,the proposed MIM-Net can effectively help the network learn more features and can be used to detect TB in CXR images,thus assisting doctors in diagnosing.