"These AI tools have a real chance to make a difference," said Eric Topol, a cardiologist and geneticist at Scripps Research in San Diego, and author of a forthcoming book on the use of the drug. automatic learning in health care, which did not participate in the research. . "But it'll take a moment."
Using neural network technology, Dr. Zhang has developed systems capable of analyzing the eyes for haemorrhages, lesions and other signs of diabetic blindness. Ideally, such systems would be a first line of defense to screen patients and identify those who need more attention.
Dr. Zhang and his colleagues have now created a system that can diagnose an even wider range of conditions, recognizing patterns in the text, not just in medical images. The new system analyzed the electronic medical records of nearly 600,000 patients at the Guangzhou Women's and Children's Medical Center, a hospital in southern China.
First, a group of qualified doctors annotated Guangzhou's records, adding labels identifying information about certain medical conditions. The system then analyzed the tagged data. When this was done and new data was presented – the symptoms of a patient determined during a physical examination – he was able to create links by himself.
When tested on untagged data, the system can compete with the performance of experienced physicians. The diagnosis of asthma was accurate to more than 90%; the accuracy of the physicians participating in the study ranged from 80 to 94%. To diagnose gastrointestinal diseases, the accuracy of the system was 87%, compared to 82 to 90% for doctors.
The experts said extensive clinical trials are now needed, not least because of the difficulty of interpreting decisions made by neural networks.
"Medicine is a slowly evolving field," said Ben Shickel, a researcher at the University of Florida, specializing in the use of in-depth learning for health care. "Nobody will just deploy one of these techniques without rigorous tests that show exactly what's going on."