Current datasets for lung computerized tomography (CT) scans are limited due to the quality and number of images, and the evaluation of these CT images are qualitative, making them difficult to standardize. In this study, Zaffino et al. provide an open-source dataset from 50 RT-PCR confirmed COVID-19 patients to augment machine learning algorithms’ translational capabilities and provide physician decision support. To assess the CT images, a radiologist assigned a clinical score, ranging from zero to five, based on the amount of ground glass opacities, consolidation, and denser tissue present in each CT scan. Zero represented 0% lung involvement, while five represented >75% lung involvement. The CT scans were segmented and filtered to define areas of interest based on voxel density. Zaffino et al. used these datasets to create a Gaussian mixture model (GMM) to detect tissue types and score each scan before validating the GMM using the clinical condition.