False-Negative Results of Real-Time Reverse-Transcriptase Polymerase Chain Reaction for Severe Acute Respiratory Syndrome Coronavirus 2: Role of Deep-Learning-Based CT Diagnosis and Insights from Two Cases

Li, Dasheng; Wang, Dawei; Dong, Jianping; Wang, Nana; Huang, He; Xu, Haiwang; Xia, Chen

The patient had been admitted to the Fever Clinic 2 weeks before because of influenza A infection as evidenced by a weakly positive nucleic acid test result. Subsequently, the patient underwent isolated medical observation before his family was diagnosed with COVID-19. During the medical observation, the nucleic acid test presented weakly positive for influenza A again, and CT showed diffuse groundglass opacities in both lungs. A deep learning (DL)-based computer-aided diagnostic system for pneumonia, which was trained with CT scans of patients with COVID-19, suggested this patient to have pneumonia, with the lesion volume accounting for 13.3% of the whole lungs ( Fig. 1) . Later, throat swab specimens from the patient were tested with rRT-PCR for SARS-CoV-2. After two consecutive negative results, a third SARS-CoV-2 rRT-PCR test confirmed the infection.