A group of Chinese scientists and Shenzhen-based tech firm, Tencent have created a model that uses artificial intelligence to quickly assess the likelihood of a patient’s Covid-19 infection turning severe or life-threatening.
Research, published in the gold-standard journal Nature, showed that nearly 7% of Covid-19 cases can deteriorate rapidly, resulting in a near 50% mortality rate.
Lung scan diagnostics can take up to 15 minutes per patient, according to South China Morning Post. The new predictor is said to establish the likelihood of deterioration within the following five, 10 and 30 days through data points.
The scientists, who include well-known Chinese infectious disease expert Dr Zhong Nanshan, credited with helping China manage the SARS outbreak in 2003, think the new model will be able to lessen the burden on medical staff by allowing staff to prioritise cases.
The model is based on the data of nearly 1 600 patients collected from various Chinese institutions. Doctors highlighted ten key areas which could provide insight into the patient’s health status. These areas include x-ray abnormalities, cancer history, age and seven other indicators.
The research and testing were conducted at Tencent AI Lab which is headed up by Zhong Nanshan. The predictor is called the ‘Calculation Tool for Early Triage of Critically Ill COVID-19 Patients using Deep Learning’. The lab made the predicter available online for all medical professionals to use.
China is not unique in its approach. IBM in the US has used its AI machine, Watson, to apply machine learning to analyse and combat the pandemic.
For the meantime, Tencent’s AI Lab will focus on Covid-based activities. The group said it would be focusing long-term on using big data and AI to prevent, control and research respiratory and thoracic diseases. Tencent has already used AI and analytics to better diagnose various types of skin cancer.
https://www.nature.com/articles/s41467-020-17280-8