NGHIÊN CỨU ĐẶC ĐIỂM HÌNH ẢNH VIÊM PHỔI COVID-19 TRÊN PHIM CẮT LỚP VI TÍNH NGỰC VÀ MỐI LIÊN QUAN VỚI MỨC ĐỘ BỆNH

Abstract

Purpose: To describe the imaging characteristics on computed tomography (CT) of lung lesions caused by COVID-19 and their relationship with disease severity. Materials and methods: A retrospective descriptive study included 160 patients with a confirmed diagnosis of COVID-19 by RT-PCR, with complete medical history information on electronic medical records and chest CT scans at the admission in COVID-19 departement of Hanoi Medical University Hospital from September-2021 to January-2023. Results: Typical COVID-19 lung lesions included ground-glass opacity (GGO - 98.1%), consolidation (67.5%), dilated vessels (39.3%), crazy-paving (31.8%) and bronchiectasis (23.8%). The distribution of COVID-19 lung lesions usually predominates in the lower lobe (77.5%), peripheral (67.5%) and posterior segment (40%) and bilateral (94.4%). Uncommon and non-specific lesions included pleural effusion (11,9%), pericardial effusion (6.3%). GGO often appear regardless of the severity of the disease (p<0.01). Meanwhile, consolidation (p<0.01), dilated vessels (p=0.033) and bronchiectasis (p<0.01) appeared to increase gradually with disease severity. The shape of opacification: round shape is dominant at mild and moderate severity (p<0.01), while at more severe levels, the shape tends to be diffuse geographic configuration (p<0.01). Conclusion: The imaging features of COVID-19 lung lesions are very typical and closely related to disease severity, so the assessment of lung lesions and their distribution contributes to increase the specificity of the diagnosis and disease severity.