Pattern recognition of high-resolution computer tomography (HRCT) chest to guide clinical management in patients with mild to moderate COVID-19

Aim: To describe the distribution of lung patterns determined by High Resolution Computed Tomography (HRCT) in COVID patients with mild and moderate lung involvement and outcomes after early identification and management with steroids and anticoagulants. Material and Methods: A cross sectional study of COVID-19 patients with mild and moderate lung involvement presenting at 5 healthcare centres in Trichy district of South TamilNadu in India. Patients underwent HRCT to assess patterns and severity of lung involvement, Inflammatory markers (LDH/Ferritin) and D-Dimer assay and clinical correlation with signs and symptoms. Patients were assessed for oxygen, steroid and anticoagulant therapy, clinical recovery or progression on follow up and details on mortality were collected. The RSNA, Fleischer Society guidelines and CORADS score was used for radiological reporting. New potential classification of patterns of percentage of lung parenchyma involvement in Covid patients is being suggested. Results: The study included 7,340 patients with suspected COVID and 3,963 (53.9%) patients had lung involvement based on HRCT. RT PCR was positive in 74.1% of the CT Positive cases. Crazy Pavement pattern was predominant (n = 2022, 51.0%) and Ground Glass Opacity (GGO) was found in 1,941 (49.0%) patients in the study. Severe lung involvement was more common in the Crazy Pavement pattern. Patients with GGO in moderate lung involvement were significantly more likely to recover faster compared to Crazy Pavement pattern (P value