4 AI systems outperforming medical experts 

Here are four recent studies in which artificial intelligence was proven to detect a health condition or make a diagnosis with greater speed or accuracy than a human expert. 

1. The artificial intelligence used by Google's DeepMind more accurately diagnosed breast cancer in women compared to six U.S. physicians, leading to fewer false positives and false negatives. In a Jan. 1 study published in Nature, the AI reduced the number of false positives, where a mammogram incorrectly diagnoses an X-ray as abnormal, by 5.7 percent among U.S. patients. Additionally, there was a 9.4 percent reduction in false negatives, which is when the cancer goes undetected.

2. A two-part approach combining AI and real-time imaging technology provided neurosurgeons with more accurate diagnoses within minutes at the bedside, allowing for more rapid treatment of potential brain tumors. Not only did the AI system achieve diagnostic accuracy of nearly 95 percent, but it was also able to identify human pathologists' misdiagnoses.

3. Six days before the CDC's Jan. 6 alert of a flu-like outbreak in China, and nine days before the World Health Organization's Jan. 9 notice, an AI-powered platform named BlueDot had already detected and sent warning of the coronavirus outbreak. BlueDot's AI algorithm analyzes global news reports, animal and plant disease networks, airline ticketing data and official announcements to predict and detect potential epidemics.

4. Google developed an AI system that detects age-related macular degeneration and demonstrates the potential to predict the disease's progression. The AI system involves two deep convolutional neural networks, one inputting raw 3D scans and the other taking a segmentation map detailing the types of tissue in the retina to predict whether the disease will develop in a patient's other eye within six months.

Copyright © 2024 Becker's Healthcare. All Rights Reserved. Privacy Policy. Cookie Policy. Linking and Reprinting Policy.

 

Featured Whitepapers

Featured Webinars