Deep learning Enables Accurate Diagnosis of Novel Coronavirus (COVID-19) with CT images

Background A novel coronavirus (COVID-19) has emerged recently as an acute respiratory syndrome. The outbreak was originally reported in Wuhan, China, but has subsequently been spread world-widely. As the COVID-19 continues to spread rapidly across the world, computed tomography (CT) has become essentially important for fast diagnoses. Thus, it is urgent to develop an accurate computer-aided method to assist clinicians to identify COVID-19-infected patients by CT images.

Materials and Methods We collected chest CT scans of 88 patients diagnosed with the COVID-19 from hospitals of two provinces in China, 101 patients infected with bacteria pneumonia, and 86 healthy persons for comparison and modeling. Based on the collected dataset, a deep learning-based CT diagnosis system (DeepPneumonia) was developed to identify patients with COVID-19.

Results The experimental results showed that our model can accurately identify the COVID-19 patients from others with an excellent AUC of 0.99 and recall (sensitivity) of 0.93. In addition, our model was capable of discriminating the COVID-19 infected patients and bacteria pneumonia-infected patients with an AUC of 0.95, recall (sensitivity) of 0.96. Moreover, our model could localize the main lesion features, especially the ground-glass opacity (GGO) that is of great help to assist doctors in diagnosis. The diagnosis for a patient could be finished in 30 seconds, and the implementation on Tianhe-2 supercompueter enables a parallel executions of thousands of tasks simultaneously. An online server is available for online diagnoses with CT images by http://biomed.nscc-gz.cn/server/Ncov2019.

Conclusions The established models can achieve a rapid and accurate identification of COVID-19 in human samples, thereby allowing identification of patients.