Chinese e-commerce giant and Amazon competitor Alibaba has trained artificial intelligence to detect coronavirus cases.
The diagnostic system should be able to identify the so-called SARS-CoV-2 (COVID-19) virus in just 20 seconds with an accuracy of 96 percent, as reported by Nikkei Asian Review.
What exactly does the AI algorithm examine?
The AI system trained by Alibaba's research team "Damo Academy" is used in a medical environment. It evaluates CT (computed tomography) scans of the lungs.
More than 5,000 cases as a basis: The researchers presented the training AI with chest (lung) images of thousands of confirmed coronavirus cases and trained them with them. According to the researchers, the AI is able to differentiate the virus from normal pneumonia (pneumonia) with an accuracy of 96 percent.
Use in clinics
According to the Nikkey Asian Review, the diagnostic system is now being used in the Qiboshan hospital in Zhengzhou and will soon be used in over 100 other clinics in several Chinese provinces.
There, it should help to identify the coronavirus faster, because it usually takes a doctor about 15 minutes to analyze a CT scan. Hundreds of images are often generated during a computed tomography scan – the AI should be able to do this within 20 seconds.
Whether comparable methods are being worked on in Europe or even used, is not known.
AI can also be helpful elsewhere. Corresponding systems control the entry into Chinese cities and can even measure the body temperature of the occupants, as reported by German.China.org.cn.
