Nanshan Zhong’s research team worked with Tencent AI Lab recently and disclosed the research results of using AI tool to predict the disease progression of patients with COVID-19. The tool can predict the probability of critical illness in 5 days, 10 days and 30 days, respectively, which is conductive to early triage of critically-ill COVID-19 patients. The research was published on the sub-journal of Nature, Nature Communications, on July 15^th, 2020.

The survival model of this research established on the basis of AI deep learning can perform an analysis of the 10 clinical characteristics of patients with COVID-19, which can help predict the progression of the disease. If this model is continued to be adopted for analysis when the patients are hospitalized, the results will be more accurate. It is conductive to monitoring the risk trend of the patients during hospitalization. The prediction tool based on this model, “Calculation Tool for Early Triage of Critically-ill COVID-19 Patients” is already released on https://aihealthcare.tencent.com/COVID19-Triage_en.html. Clinicians and medical staff can also visit its Wechat Mini Program to use the tool.

Medical staff only need to enter the clinical characteristics of the patient, and Calculation Tool for Early Triage of Critically-ill COVID-19 Patients can provide the probability that the patient will develop to a critical condition within 5, 10, and 30 days, and then perform early triage of the patient, which is extremely useful and highly economical for the management of COVID-19.

Meanwhile, the source of this research result is also open to the world through Github to support the global fight against COVID-19. The query link of open source projects on Github is https://github.com/cojocchen/covid19_critically_ill.