CXR-Net: A Multitask Deep Learning Network for Explainable and Accurate Diagnosis of COVID-19 Pneumonia from Chest X-ray Images

Zhang, Xin ORCID: https://orcid.org/0000-0001-7844-593X, Han, Liangxiu ORCID: https://orcid.org/0000-0003-2491-7473, Sobeih, Tam, Han, Lianghao ORCID: https://orcid.org/0000-0003-2491-7473, Dempsey, Nina ORCID: https://orcid.org/0000-0003-2013-6207, Lechareas, Symeon, Tridente, Ascanio, Chen, Haoming, White, Stephen ORCID: https://orcid.org/0000-0003-0090-6358 and Zhang, Daoqiang (2023) CXR-Net: A Multitask Deep Learning Network for Explainable and Accurate Diagnosis of COVID-19 Pneumonia from Chest X-ray Images. IEEE journal of biomedical and health informatics, 27 (2). pp. 980-991. ISSN 2168-2194

Preview

Accepted Version Available under License In Copyright. Download (897kB) | Preview

Abstract

Accurate and rapid detection of COVID-19 pneumonia is crucial for optimal patient treatment. Chest X-Ray (CXR) is the first-line imaging technique for COVID-19 pneumonia diagnosis as it is fast, cheap and easily accessible. Currently, many deep learning (DL) models have been proposed to detect COVID-19 pneumonia from CXR images. Unfortunately, these deep classifiers lack the transparency in interpreting findings, which may limit their applications in clinical practice. The existing explanation methods produce either too noisy or imprecise results, and hence are unsuitable for diagnostic purposes. In this work, we propose a novel explainable CXR deep neural Network (CXR-Net) for accurate COVID-19 pneumonia detection with an enhanced pixel-level visual explanation using CXR images. An Encoder-Decoder-Encoder architecture is proposed, in which an extra encoder is added after the encoder-decoder structure to ensure the model can be trained on category samples. The method has been evaluated on real world CXR datasets from both public and private sources, including healthy, bacterial pneumonia, viral pneumonia and COVID-19 pneumonia cases. The results demonstrate that the proposed method can achieve a satisfactory accuracy and provide fine-resolution activation maps for visual explanation in the lung disease detection. The Average Accuracy, Sensitivity, Specificity, PPV and F1-score of models in the COVID-19 pneumonia detection reach 0.992, 0.998, 0.985 and 0.989, respectively. Compared to current state-of-the-art visual explanation methods, the proposed method can provide more detailed, high-resolution, visual explanation for the classification results. It can be deployed in various computing environments, including cloud, CPU and GPU environments. It has a great potential to be used in clinical practice for COVID-19 pneumonia diagnosis.