An interpretable multi-task system for clinically applicable COVID-19 diagnosis using CXR

Abstract

BACKGROUND: With the emergence of continuously mutating variants of coronavirus, it is urgent to develop a deep learning model for automatic COVID-19 diagnosis at early stages from chest X-ray images. Since laboratory testing is time-consuming and requires trained laboratory personal, diagnosis using chest X-ray (CXR) is a befitting option. OBJECTIVE: In this study, we proposed an interpretable multi-task system for automatic lung detection and COVID-19 screening in chest X-rays to find an alternate method of testing which are reliable, fast and easily accessible, and able to generate interpretable predictions that are strongly correlated with radiological findings. METHODS: The proposed system consists of image preprocessing and an unsupervised machine learning (UML) algorithm for lung region detection, as well as a truncated CNN model based on deep transfer learning (DTL) to classify chest X-rays into three classes of COVID-19, pneumonia, and normal. The Grad-CAM technique was applied to create class-specific heatmap images in order to establish trust in the medical AI system. RESULTS: Experiments were performed with 15,884 frontal CXR images to show that the proposed system achieves an accuracy of 91.94% in a test dataset with 2,680 images including a sensitivity of 94.48% on COVID-19 cases, a specificity of 88.46% on normal cases, and a precision of 88.01% on pneumonia cases. Our system also produced state-of-the-art outcomes with a sensitivity of 97.40% on public test data and 88.23% on a previously unseen clinical data (1,000 cases) for binary classification of COVID-19-positive and COVID-19-negative films. CONCLUSION: Our automatic computerized evaluation for grading lung infections exhibited sensitivity comparable to that of radiologist interpretation in clinical applicability. Therefore, the proposed solution can be used as one element of patient evaluation along with gold-standard clinical and laboratory testing.

COVID-19 Complications: Oxidative Stress, Inflammation, and Mitochondrial and Endothelial Dysfunction

“…The entry of the SARS-CoV-2 virus into the respiratory system and the subsequent initiation of a strong inflammatory response lead to the breakdown of the alveolar-capillary barrier [37]. In the acute stage of infection, ARDS causes lung damage that includes the formation of hyaline membranes in the alveoli, followed by interstitial expansion, edema, and fibroblast proliferation with typical pathological changes characterized by diffuse alveolar parenchymal damage [38]. The accumulation of protein-rich fluid in the alveolar and interstitial spaces causes pulmonary surfactant inhibition [35].…”

Automatic Detection of COVID-19 in the Lungs X-ray Images using Pre-trained Deep Learning Model CNN

“…Artificial Intelligence (AI) models, medical experience, and image research may have the energy to produce a beneficial, lasting effect on people's lives in a relatively small-time span. The analysis by computer and medical image analysis includes image recovery, image processing, image study, and image-based envision [10]. The medicalimage analysis has heightened to incorporate computer vision, pattern-identification, image-separation, and machine learning (ML) in many dimensions [ 12].…”

Automated Quantification of Pneumonia Infected Volume in Lung CT Images: A Comparison with Subjective Assessment of Radiologists

“…Recently, in developing CAD schemes of medical images, deep learning (DL) models have been well recognized and widely used to perform the tasks of segmenting the disease-infected regions of interest (ROIs) [ 7 , 8 ] and detecting or classifying diseases using the automatically extracted image features [ 9 , 10 ]. In using COVID-19 image datasets to develop CAD schemes, most of the previous studies focused on developing DL models to detect COVID-19 cases or classify between the COVID-19 and normal or other types of pneumonia cases [ 11 , 12 , 13 , 14 ]. Although many previous studies reported the extremely high accuracy of using DL models to detect and/or classify the COVID-19 infected cases (i.e., ranging from 90–100% accuracy [ 15 ]), no previous DL model is robust and clinically acceptable due to training bias and a “black-box” type approach [ 16 ].…”