AI leads drive for Healthy China

The in-depth integration of artificial intelligence (AI) technology into the medical industry in recent years is creating more possibilities for people’s health and medical services.

Smart watches are used to monitor ECG and atrial fibrillation; video meetings are launched to conduct multi-discipline consultation for patients in remote areas; and smart phones can help people with Parkinson’s disease complete daily motion function assessment.

“The development of smart health services and AI technology has to a large extent facilitated doctors and brought convenience to health workers,” said Zhou Yujie, vice president of Beijing Anzhen Hospital and expert in cardiovascular diseases. The data and models gained will be included in more human-machine diagnosis and treatment schemes in the future and serve as a bedrock for smart medical industry, he added.

Ping An Technology, a technology solution expert of China’s leading insurer Ping An Group, has developed a smart COVID-19 image analysis system which is able to generate CT image results for COVID-19 patients in just 15 seconds, while a doctor normally has to spend 5 to 15 minutes on this process.

The system, since Feb. 21 this year, has been introduced to 1,500 medical institutions across China, including Hubei, former epicenter of COVID-19, greatly improving the working efficiency of the doctors. During China’s battle against the disease, the system spotted over 20,000 suspected COVID-19 cases and helped doctors conduct further screening work. It has analyzed more than 4 million CT images in total.

CT images are of vital importance in the diagnosis of COVID-19, and the surging number of patients brought huge pressure for radiologists, Xie Guotong, chief healthcare scientist of the Ping An Group told the People’s Daily. “They were in urgent demand of an assistant that is able to analyze images with high accuracy and speed,” Xie added.

According to Zhou, Beijing Anzhen Hospital has to perform two to three hundred CT angiography operations every day. “It took over 4 days from examinations to report printing, because all the procedures were manually operated. Thanks to an AI-assisted cardiovascular diagnosis system developed by Shukun Technology, our efficiency has been doubled,” Zhou introduced.

AI software is able to precisely detect hemadostenosis in CT examination for coronary heart disease. “We used to watch with our eyes, and stents would be inserted when the hemadostenosis reaches certain points. But now whether we use stents depends on how the patients’ blood vessels function, and the AI software can rate the hemadostenosis and blood vessel function of the patients in 5 minutes,” Zhou introduced. “As long as the vessels function well, we won’t insert stents,” he added, calling it a milestone progress for the treatment of cardiovascular diseases.

Zhou told the People’s Daily that doctors used to manually record the medical history of patients, so it was hard to conduct structured analysis. Now, with AI and big data, both the management of the patients and the analysis of their states can be structuralized. For instance, doctors are able to acquire the characteristics of the cardiovascular disease-susceptible group above the age of 40, as well as the potential signals of pathological changes at different ages. “It helps a lot with the prevention and diagnosis of diseases,” Zhou added.

AI imaging and AI-based auxiliary diagnosis are widely applied nowadays, in which AI aids the doctors in diagnosing based on the images, case reports and in vitro diagnosis of the patients. According to incomplete statistics, AI has offered auxiliary diagnosis for a total of over 10 million patients with lung or cardiovascular diseases.

With AI technology, a medical image reading platform developed by Chinese tech giant Tencent can help doctors identify and predict the risks of over 700 diseases, thus largely improving the accuracy and efficiency of diagnosis.

In addition, AI also facilitates post-diagnosis management. It is able to assess the situation of dyskinesia patients through videos in a more efficient and precise manner, classify the patients, and make more targeted treatment plans. Besides, doctors can also interact with patients through mobile phones for better management of the patients.