An Artificial intelligence system developed by Google is better at spotting breast cancer than radiologists, suggests a study published in the science journal Nature.
The scientists used mammogram images of women from the UK and the US. Six radiologists were pitted against the Google AI and Deep Mind system for diagnosing incidents of breast cancer in 500 random mammogram images. The computer outperformed the doctors on an average of 11.5 per cent.
They also compared the performance with the actual results from a set of 25,856 mammograms in the UK and 3,097 from the US; here too, the AI performed better. The computer software, developed by Alphabet’s AI unit Deep Mind and Google Health, reduced the number of false negatives (ruling cancer out in cases that do have it) by 9.4 per cent in the US and 2.7 per cent in the UK. And the number of false positives (wrongly diagnosing cancer where there is none) fell by 5.7 per cent in the U.S. and 1.2 per cent in the UK, the research found. The AI system scanned nearly 29,000 images.
The research team comprised of researchers from Google Health, Northwestern Medicine and Imperial College London.
“These results highlight the significant role that AI could play in the future of cancer care. Embracing technology like this may help improve the way we diagnose cancer in the years to come,” said UK Cancer Research chief Executive Michelle Mitchell in a statement.
In a double-reading simulated process for detecting cancer ( double screening process is used in the UK ), the AI system maintained its performance. It reduced the workload of the second reader by 88 per cent.
AI Spots Cancer Better
Breast cancer is the second leading cause of death among women in the world and affects one in eight women. Half of the women who get screened for breast cancer over a 10-year period get a false-positive result, according to the American Cancer Society. The use of AI in screening, which is the most common method of finding out if a person suffers from the ailment, can help in reducing false diagnosis by a considerable margin.
Connie Lehman, chief of the breast imaging department at Harvard’s Massachusetts General Hospital, said to The Independent that the results matched earlier findings on AI’s role in improving mammogram based diagnosis of cancer, including her research.
Computer-aided detection (CAD)of cancer is not new and has been in operation since the past few decades. But CADS have not been too useful in actual clinical practice. This was because the systems were subordinate to the human detection practice rather than using intuitive data-based results. The present AI uses 1000s images to spot the anomalies.
AI aided cancer detection has the potential to “exceed human capacity to identify subtle cues that the human eye and brain aren’t able to perceive,” Dr Lehman said.