A Deep Learning System to Screen Novel Coronavirus Disease 2019 Pneumonia

摘要: We found that the real time reverse transcription-polymerase chain reaction (RT-PCR) detection of viral RNA from sputum or nasopharyngeal swab has a relatively low positive rate in early stage to determine COVID-19 (named by World Health Organization). The manifestations computed tomography (CT) imaging had their own characteristics, which are different other types pneumonia, such as Influenza-A pneumonia. Therefore, clinical doctors call for another diagnostic criteria this new type pneumonia soon possible.This study aimed establish an screening model distinguish and healthy cases with pulmonary CT images using deep learning techniques. candidate infection regions were first segmented out 3-dimensional image set. These separated then categorized into COVID-19, irrelevant groups, together corresponding confidence scores location-attention classification model. Finally total score case calculated Noisy-or Bayesian function.The experiments result benchmark dataset showed overall accuracy was 86.7 % perspective whole.The models established effective patients demonstrated be promising supplementary method frontline doctors.