Estimating COVID-19 Pneumonia Extent and Severity From Chest Computed Tomography

Area of Science:

• Radiology
• Pulmonology
• Medical Imaging

Background:

• Computed tomography (CT) is used to assess COVID-19 pneumonia extent via the ratio of abnormal pulmonary opacities (PO) volume to estimated lung volume (CT_LV).
• CT-estimated lung weight (CT_LW) also correlates with pneumonia severity, but CT_LV and CT_LW are influenced by demographic and anthropometric factors.

Purpose of the Study:

• To estimate COVID-19 pneumonia extent and severity by adjusting abnormal PO volume and weight to predicted CT lung volume (pCT_LV) and predicted CT lung weight (pCT_LW).
• To evaluate the association of these adjusted metrics with clinical and radiological outcomes.

Main Methods:

• Retrospective analysis of chest CT scans from 103 COVID-19 patients and 86 healthy controls.
• Development of predictive equations for pCT_LV and pCT_LW in controls.
• Definition of pneumonia extent and severity as percentages of pCT_LV and pCT_LW, respectively, with Z-score assessment against controls. ROC analysis was used for differential diagnosis.

Main Results:

• Predictive equations for CT_LV (sex, height) and CT_LW (age, sex, height) were established in controls.
• Cutoffs for pneumonia extent (20%) and severity (50%) showed good diagnostic ability (AUC > 0.90).
• Adjusted pneumonia extent correlated better with pCT_LV and pCT_LW (r=0.85). Severe/diffuse pneumonia correlated with higher CRP, mortality, ICU stay, and ventilation needs. Severity correlated with age and CRP.

Conclusions:

• The proposed method for estimating COVID-19 pneumonia extent and severity using predicted lung parameters offers a potentially valuable tool.
• This approach may enhance clinical and radiological patient stratification for improved management and outcomes.