Attention-Guided Partial Domain Adaptation for Automated Pneumonia Diagnosis From Chest X-Ray Images

• Published in: IEEE journal of biomedical and health informatics 2023-12, Vol.27 (12), p.5848-5859
• Main Authors: Liu, Wentao, Ni, Zhiwei, Chen, Qian, Ni, Liping
• Format: Article
• Language: English
• Subjects: Adaptation; adversarial learning; Artificial neural networks; Biomedical imaging; Chest; COVID-19; COVID-19 - diagnostic imaging; COVID-19 Testing; Datasets; domain adaptation; Health Status; Heterogeneity; Humans; Lesions; Lung; Minority & ethnic groups; negative transfer; Neural networks; Pneumonia; Pneumonia - diagnostic imaging; Pneumonia diagnosis; Pulmonary diseases; Semantics; Task analysis; Thorax; X-ray imaging; X-Rays
• ISSN: 2168-2194; 2168-2208; 2168-2208
• DOI: 10.1109/JBHI.2023.3313886

Deep neural networks (DNN) supported by multicenter large-scale Chest X-Ray (CXR) datasets can efficiently perform tasks such as disease identification, lesion segmentation, and report generation. However, the non-ignorable inter-domain heterogeneity caused by different equipment, ethnic groups, and scanning protocols may lead to dramatic degradation in model performance. Unsupervised domain adaptation (UDA) methods help alleviate the cross-domain discrepancy for subsequent analysis. Nevertheless, they may be prone to: 1) spatial negative transfer: misaligning non-transferable regions which have inadequate knowledge, and 2) semantic negative transfer: failing to extend to scenarios where the label spaces of the source and target domain are partially shared. In this work, we propose a classification-based framework named attention-guided partial domain adaptation (AGPDA) network for overcoming these two negative transfer challenges. AGPDA is composed of two key modules: 1) a region attention discrimination block (RADB) to generate fine-grained attention value via lightweight region-wise multi-adversarial networks. 2) a residual feature recalibration block (RFRB) trained with class-weighted maximum mean discrepancy (MMD) loss for down-weighing the irrelevant source samples. Extensive experiments on two publicly available CXR datasets containing a total of 8598 pneumonia (viral, bacterial, and COVID-19) cases, 7163 non-pneumonia or healthy cases, demonstrate the superior performance of our AGPDA. Especially on three partial transfer tasks, AGPDA significantly increases the accuracy, sensitivity, and F1 score by 4.35%, 4.05%, and 1.78% compared to recently strong baselines.