A fully automatic deep learning system for COVID-19 diagnostic and prognostic analysis

TLDR

COVID‑19 has spread worldwide, straining medical resources and underscoring the need for rapid diagnosis and identification of high‑risk patients. The study proposes a fully automatic deep‑learning system that performs diagnostic and prognostic analysis of COVID‑19 using routine computed tomography. Using 5,372 CT scans, the system was pre‑trained on 4,106 cases to learn lung features and then trained and externally validated on 1,266 patients (924 COVID‑19, 342 other pneumonia) from six regions. In four external validation sets, the system achieved AUCs of 0.87–0.88 for COVID‑19 versus other pneumonia and 0.86 versus viral pneumonia, accurately stratified patients into high‑ and low‑risk groups with significantly different hospital stays, and automatically highlighted abnormal areas consistent with radiological findings, offering a fast screening tool for resource optimization.

Abstract

Coronavirus disease 2019 (COVID-19) has spread globally, and medical resources become insufficient in many regions. Fast diagnosis of COVID-19 and finding high-risk patients with worse prognosis for early prevention and medical resource optimisation is important. Here, we proposed a fully automatic deep learning system for COVID-19 diagnostic and prognostic analysis by routinely used computed tomography. We retrospectively collected 5372 patients with computed tomography images from seven cities or provinces. Firstly, 4106 patients with computed tomography images were used to pre-train the deep learning system, making it learn lung features. Following this, 1266 patients (924 with COVID-19 (471 had follow-up for >5 days) and 342 with other pneumonia) from six cities or provinces were enrolled to train and externally validate the performance of the deep learning system. In the four external validation sets, the deep learning system achieved good performance in identifying COVID-19 from other pneumonia (AUC 0.87 and 0.88, respectively) and viral pneumonia (AUC 0.86). Moreover, the deep learning system succeeded to stratify patients into high- and low-risk groups whose hospital-stay time had significant difference (p=0.013 and p=0.014, respectively). Without human assistance, the deep learning system automatically focused on abnormal areas that showed consistent characteristics with reported radiological findings. Deep learning provides a convenient tool for fast screening of COVID-19 and identifying potential high-risk patients, which may be helpful for medical resource optimisation and early prevention before patients show severe symptoms.