AnalyzeDirect Aaron So et al. Relationship between lung injury extent and phenotype manifested in non-contrast CT and cardiac injury during acute stage of COVID-19 Laboratory Investigation.
Publication Date
December 25, 2021
How Analyze was Used
“Analysis of NCCT chest images was performed with the Analyze 14.0 software (AnalyzeDirect Inc., Overland Park, Kansas). From each set of NCCT images, the left and right lungs across all tomographic slices were delineated with the automatic contour detection function in Analyze. The detected contour of each lung was then modified manually if needed to minimize the coverage of the trachea, bronchi and large pulmonary blood vessels. Thereafter, the CT number in each voxel in each delineated lung region and the corresponding image voxel dimensions (length, width and depth) were exported using the histogram function in Analyze. Next, the volume of each lung was calculated as the number of image voxel covering the lung multiplied by the image voxel dimensions.
Normal lung parenchyma was defined as the image voxel within the lungs that had a CT number between −950 and −700 HU. Abnormal lung parenchyma was further classified into two categories according to their differences in tissue characteristics[12]: moderately abnormal parenchyma, which had a low elastance (high compliance) and manifested as ground glass opacification, had a mean CT number between −600 to −150 HU; severely abnormal parenchyma, which had a high elastance (low compliance) and manifested as consolidation, had a mean CT number between −149 to 150 HU (Fig. 1). The relative volume of normal lung parenchyma (RVNP), moderately abnormal lung parenchyma (RVMAP) and severely abnormal lung parenchyma (RVSAP) with respect to the total lung volume (sum of the left and right lung volumes) were calculated accordingly.”
Keywords
COVID-19, SARS-CoV-2, Acute cardiac injury, Non-contrast computed tomography (NCCT), Inflammation
Author Affiliation(s)
Lawson Health Research Institute, London, Ontario, Canada
