Artificial Intelligence Defeated Human Doctors In Contest To Correctly Diagnose Tumors

"I hope through this competition, doctors can experience the power of artificial intelligence."

A new technology is putting China's top doctors to the test. The BioMind AI system recently defeated a team of high-ranking doctors when asked to diagnose brain tumors and predict hematoma expansions as part of a competition, according to Xinhua News Agency.

The Chinese government's state-run press agency reported that the Beijing Tiantan Hospital's Artificial Intelligence Research Center for Neurological Disorders developed the medically savvy artificial intelligence system. Per the agency, when compared to the capabilities of human professionals, the Al system was faster and more accurate. . 


During the competition, the AI system diagnosed brain tumors with 87 percent accuracy. Human medical professionals were correct only 66 percent of the time. Similar disparities emerged when the system and the doctors were tasked with predicting brain hematoma expansion.

The BioMind Al system was also faster than the doctors. The technology took 15 minutes to diagnose 225 cases, while its human competitors took 30 minutes. Per Xinhua, the system was fed a series of images from Beijing Tiantan Hospital's archives that taught it to detect certain diseases. Over the years, the Al system has made great progress

Softbank Pepper robot provide assistance to customers in automation fair Turin Italy April 18 2018. Antonello Marangi /

However, researchers emphasized that the purpose of the system is to support the work of medical professionals and not to replace them.

"I hope through this competition, doctors can experience the power of artificial intelligence," Wang Yongjun, executive vice president of the Tiantan Hospital, told Xinhua. "This is especially true for some doctors who are skeptical about artificial intelligence. I hope they can further understand AI and eliminate their fears toward it."

The Al system is still in its early development phases, but similar gains are being made in applying machine learning to medical diagnoses all over the world. Earlier this month, The MIT Technology Review's Physics arXiv Blog noted that two German scientists, Nils Strodthoff and Claas Strodthoff, had designed a neural network that is able to detect heart attacks at a level similar to human cardiologists.  

"Only with bigger data sets can clinicians be sure the algorithms will be accurate in the wide range of chaotic environments that doctors work in," the blog noted of algorithms that, like the BioMind AI system, are powered by large data sets. "But the potential is huge. Machines can relieve some of medicine's tedium and complexity for human doctors—and do it without tiring."

Cover image via Shutterstock / Sdecoret.


Subscribe to our newsletter and get the latest news and exclusive updates.