分子組織学および医学生理学ジャーナル

Application of Deep Learning for Whole-Lung and Lung-Lesion Quantification in Computerized Tomography Despite Inconsistent Ground Truth

Abstract

Devashish Nath

Computed Tomography (CT) imaging is a crucial tool for diagnosing, characterizing, prognosticating and monitoring disease progression in patients infected with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). However, to evaluate lung abnormalities in a consistent and reliable manner, accurate segmentation and quantification of both the entire lung and lung lesions (abnormalities) in chest CT images of COVID-19 patients is necessary. Unfortunately, manual segmentation and quantification of a large dataset can be time-consuming and have low inter- and intra-observer agreement, even for experienced radiologists.