RESUMO
ABSTRACT: The COVID-19 pandemic has challenged institutions' diagnostic processes worldwide. The aim of this study was to assess the feasibility of an artificial intelligence (AI)-based software tool that automatically evaluates chest computed tomography for findings of suspected COVID-19.Two groups were retrospectively evaluated for COVID-19-associated ground glass opacities of the lungs (group A: real-time polymerase chain reaction positive COVID patients, nâ=â108; group B: asymptomatic pre-operative group, nâ=â88). The performance of an AI-based software assessment tool for detection of COVID-associated abnormalities was compared with human evaluation based on COVID-19 reporting and data system (CO-RADS) scores performed by 3 readers.All evaluated variables of the AI-based assessment showed significant differences between the 2 groups (Pâ<â.01). The inter-reader reliability of CO-RADS scoring was 0.87. The CO-RADS scores were substantially higher in group A (mean 4.28) than group B (mean 1.50). The difference between CO-RADS scoring and AI assessment was statistically significant for all variables but showed good correlation with the clinical context of the CO-RADS score. AI allowed to predict COVID positive cases with an accuracy of 0.94.The evaluated AI-based algorithm detects COVID-19-associated findings with high sensitivity and may support radiologic workflows during the pandemic.
Assuntos
Inteligência Artificial/normas , COVID-19/diagnóstico , Pulmão/diagnóstico por imagem , Idoso , Idoso de 80 Anos ou mais , COVID-19/epidemiologia , Teste de Ácido Nucleico para COVID-19/normas , Estudos de Viabilidade , Feminino , Humanos , Pulmão/patologia , Masculino , Pessoa de Meia-Idade , Pandemias , Estudos Retrospectivos , SARS-CoV-2 , Tomografia Computadorizada por Raios XRESUMO
BACKGROUND: The COVID-19 pandemic challenges the recommendations for patients' preoperative assessment for preventing severe acute respiratory syndrome coronavirus type 2 transmission and COVID-19-associated postoperative complications and morbidities. PURPOSE: To evaluate the contribution of chest computed tomography for preoperatively assessing patients who are not suspected of being infected with COVID-19 at the time of referral. METHODS: Candidates for emergency surgery screened via chest computed tomography from 8 to 27 April 2020 were retrospectively evaluated. Computed tomography images were analysed for the presence of COVID-19-associated intrapulmonary changes. When applicable, laboratory and recorded clinical symptoms were extracted. RESULTS: Eighty-eight patients underwent preoperative chest computed tomography; 24% were rated as moderately suspicious and 11% as highly suspicious on computed tomography. Subsequent reverse transcription polymerase chain reaction (RT-PCR) was performed for seven patients, all of whom tested negative for COVID-19. Seven patients showed COVID-19-associated clinical symptoms, and most were classified as being mildly to moderately severe as per the clinical classification grading system. Only one case was severe. Four cases underwent RT-PCR with negative results. CONCLUSION: In a cohort without clinical suspicion of COVID-19 infection upon referral, preoperative computed tomography during the COVID-19 pandemic can yield a high suspicion of infection, even if the patient lacks clinical symptoms and is RT-PCR-negative. No recommendations can be made based on our results but contribute to the debate.