RESUMEN
A critical step in effective care and treatment planning for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the cause for the coronavirus disease 2019 (COVID-19) pandemic, is the assessment of the severity of disease progression. Chest x-rays (CXRs) are often used to assess SARS-CoV-2 severity, with two important assessment metrics being extent of lung involvement and degree of opacity. In this proof-of-concept study, we assess the feasibility of computer-aided scoring of CXRs of SARS-CoV-2 lung disease severity using a deep learning system. Data consisted of 396 CXRs from SARS-CoV-2 positive patient cases. Geographic extent and opacity extent were scored by two board-certified expert chest radiologists (with 20+ years of experience) and a 2nd-year radiology resident. The deep neural networks used in this study, which we name COVID-Net S, are based on a COVID-Net network architecture. 100 versions of the network were independently learned (50 to perform geographic extent scoring and 50 to perform opacity extent scoring) using random subsets of CXRs from the study, and we evaluated the networks using stratified Monte Carlo cross-validation experiments. The COVID-Net S deep neural networks yielded R[Formula: see text] of [Formula: see text] and [Formula: see text] between predicted scores and radiologist scores for geographic extent and opacity extent, respectively, in stratified Monte Carlo cross-validation experiments. The best performing COVID-Net S networks achieved R[Formula: see text] of 0.739 and 0.741 between predicted scores and radiologist scores for geographic extent and opacity extent, respectively. The results are promising and suggest that the use of deep neural networks on CXRs could be an effective tool for computer-aided assessment of SARS-CoV-2 lung disease severity, although additional studies are needed before adoption for routine clinical use.
RESUMEN
AIM: To evaluate whether portable chest radiography (CXR) scores are associated with coronavirus disease 2019 (COVID-19) status and various clinical outcomes. MATERIALS AND METHODS: This retrospective study included 500 initial CXR from COVID-19-suspected patients. Each CXR was scored based on geographic extent and degree of opacity as indicators of disease severity. COVID-19 status and clinical outcomes including intensive care unit (ICU) admission, mechanical ventilation, mortality, length of hospitalisation, and duration on ventilator were collected. Multivariable logistic regression analysis was performed to evaluate the relationship between CXR scores and COVID-19 status, CXR scores and clinical outcomes, adjusted for code status, age, gender and co-morbidities. RESULTS: The interclass correlation coefficients amongst raters were 0.94 and 0.90 for the extent score and opacity score, respectively. CXR scores were significantly (p < 0.01) associated with COVID-19 positivity (odd ratio [OR] = 1.49; 95% confidence interval [CI]: 1.27 - 1.75 for extent score and OR = 1.75; 95% CI: 1.42 - 2.15 for opacity score), ICU admission (OR = 1.19; 95% CI: 1.09 - 1.31 for extent score and OR = 1.26; 95% CI: 1.10 - 1.44 for opacity score), and invasive mechanical ventilation (OR = 1.22; 95% CI: 1.11 - 1.35 for geographic score and OR = 1.21; 95% CI: 1.05 - 1.38 for opacity score). CXR scores were not significantly different between survivors and non-survivors after adjusting for code status (p>0.05). CXR scores were not associated with length of hospitalisation or duration on ventilation (p>0.05). CONCLUSIONS: Initial CXR scores have prognostic value and are associated with COVID-19 positivity, ICU admission, and mechanical ventilation.
