Beijing, China - June 30, 2018 (CCTV - No access Chinese mainland)
1. AI-versus-human competition site
2. Various of doctors reading scans of neuroimaging
3. Various of AI diagnosis system
4. Various of scans of neuroimaging
5. SOUNDBITE (Chinese) Gao Peiyi, neuroradiologist, Beijing Tiantan Hospital:
"It does not memorize the tumors but extract features [of them] and study them. When a new tumor is given to it, it will compare the new one with those features and see which type it belongs to. And then make the diagnosis."
6. Various of doctors reading scans of neuroimaging
7. Various of scans of neuroimaging
8. Doctor writing down diagnosis result
9. Doctors discussing results
10. Various of AI diagnosis system working
11. Various of diagnosis results of AI
12. SOUNDBITE (Chinese) Wang Yongjun, executive vice president, Beijing Tiantan Hospital (starting with shot 11):
"Tireless and stable, this tool is capable of bringing an effect to clinical applications no other tools are able to. So I hope through this competition, more clinicians and people can further understand AI."
13. Various of judges
14. Various of competition result
15. People posing for photos on stage

An artificial intelligence (AI) system scored 2:0 against elite human physicians Saturday in two rounds of competitions in diagnosing brain tumors and predicting hematoma expansion in Beijing.

In the first round of the competition, the BioMind AI system, developed by the Artificial Intelligence Research Center for Neurological Disorders at the Beijing Tiantan Hospital and a research team from the Capital Medical University, made correct diagnoses in 87 percent of 225 cases in about 15 minutes, while a team of 15 senior doctors only achieved 66-percent accuracy.

According to its developers, the AI-assisted diagnosis system studied images of nervous system-related diseases that the Beijing Tiantan Hospital has archived over the past 10 years, making it capable of diagnosing common neurological diseases such as meningioma and glioma with an accuracy rate of over 90 percent, comparable to that of a senior doctor.

"It does not memorize the tumors but extract features [of them] and study them. When a new tumor is given to it, it will compare the new one with those features and see which type it belongs to. And then make the diagnosis," said Gao Peiyi, a neuroradiologist of Beijing Tiantan Hospital.

In the second round of the completion, the AI gave correct predictions in 83 percent of brain hematoma expansion cases in three minutes, outperforming the 63-percent accuracy among a group of physicians from renowned hospitals across the country who spent more than 20 minutes.

"Tireless and stable, this tool is capable of bringing an effect to clinical applications no other tools are able to. So I hope through this competition, more clinicians and people can further understand AI," said Wang Yongjun, executive vice president of the Beijing Tiantan Hospital.

Experts said that if the AI system is widely applied in the clinic, more doctors will be freed from the heavy workload and devote themselves to more innovative endeavor. The AI system can also help hospitals to improve the diagnostic efficiency and accuracy.