Enhancing Multi-disease Diagnosis of Chest X-rays with Advanced Deep-learning Networks in Real-world Data

Abstract

The current artificial intelligence (AI) models are still insufficient in multi-disease diagnosis for real-world data, which always present a long-tail distribution. To tackle this issue, a long-tail public dataset, “ChestX-ray14,” which involved fourteen (14) disease labels, was randomly divided into the train, validation, and test sets with ratios of 0.7, 0.1, and 0.2. Two pretrained state-of-the-art networks, EfficientNet-b5 and CoAtNet-0-rw, were chosen as the backbones. After the fully-connected layer, a final layer of 14 sigmoid activation units was added to output each disease’s diagnosis. To achieve better adaptive learning, a novel loss ( L ours ) was designed, which coalesced reweighting and tail sample focus. For comparison, a pretrained ResNet50 network with weighted binary cross-entropy loss ( L WBCE ) was used as a baseline, which showed the best performance in a previous study. The overall and individual areas under the receiver operating curve (AUROC) for each disease label were evaluated and compared among different models. Group-score-weighted class activation mapping (Group-CAM) is applied for visual interpretations. As a result, the pretrained CoAtNet-0-rw + L ours showed the best overall AUROC of 0.842, significantly higher than ResNet50 + L WBCE (AUROC: 0.811, p = 0.037). Group-CAM presented that the model could pay the proper attention to lesions for most disease labels (e.g., atelectasis, edema, effusion) but wrong attention for the other labels, such as pneumothorax; meanwhile, mislabeling of the dataset was found. Overall, this study presented an advanced AI diagnostic model achieving a significant improvement in the multi-disease diagnosis of chest X-rays, particularly in real-world data with challenging long-tail distributions.

AI-Based Model Design for Prediction of COPD Grade from Chest X-Ray Images: A Model Proposal (COPD-GradeNet)

“…In recent years, artificial intelligence methods, especially deep convolutional neural networks (DCNNs), have been widely used to enable clinicians to automate tasks such as the classification of COVID-19 from CT [10] and chest X-ray images [11][12][13], determination of the severity of COVID-19 [14], diagnosis of proximal femur fracture from MR images [15], early detection of pathological changes in bone microstructures [16], disease diagnosis using laboratory test results [17], determination of the effectiveness of Shapley value in identifying low-quality and valuable data for pneumonia detection [18], and localization of common chest diseases on chest X-rays [19], and improving the performance of radiologists in breast cancer screening [20]. With the proliferation of deep learning models, the field of medical image processing has garnered widespread interest, particularly in radiology [21][22][23].…”