Google AI System Can Detect Early Signs Of Breast Cancer Signs Better Than Radiologists: Study

We’ve already witnessed examples of Artificial Intelligence being a boon in the healthcare industry. There’s AI-assisted robotic surgery, AI in the analysis of mammogram and radiology, and precision medicine among others. And there’s another new study which reveals that the Google Artificial Intelligence system is as efficient as the expert radiologists in analysis of women with breast cancer based on screening mammograms. It also helped reduce the number of false-positive results by 5.7% in the U.S.-based group and by 1.2% in the British-based group, reports Reuters. The number of false negatives were also reduced by 9.4% in the U.S. group, and by 2.7% in the British group.

SEE ALSO: Chrome’s New Feature Uses AI To Describe Images For Blind And Low-Vision Users

The study has been published in the journal Nature and reveals that AI has the potential to improve the accuracy of screening for breast cancer which affects one in eight women globally. According to Mozziyar Etemadi, co-author of the study, Northwestern Medicine in Chicago, the study, “represent a major advance in the potential for the early detection of breast cancer”. Other researchers who were part of the study are from the Imperial College London and Britain’s National Health Service.

Researchers trained the AI system to accurately detect breast cancer on over a thousand mammograms. They also compared AI’s performance with the results of 25,856 mammograms in the United Kingdom and 3,097 from the United States. The findings of the study had been developed with Alphabet Inc’s DeepMind AI unit which integrated with Google Health in September 2019.

Researchers conducted a test where Google’s AI system was pitted against six radiologists where the AI managed to outperform all six of them at accurately detecting breast cancer. According to Connie Lehman, chief of breast imaging department, Harvard Massachusetts General Hospital, “the results (of the study) are in line with findings from several groups using AI to improve cancer detection in mammograms, including her own work”.

Lehman also pointed out the difference between current CAD (Computer-aided diagnosis) programs and AI systems, stating that while the current CAD programs are trained to spot things that human radiologists can see, AI can learn to spot the signs of cancer depending on thousands of mammogram results. This, in turn, can help AI “exceed human capacity” when it comes to identifying certain hints or signs that can’t be perceived by humans.

However, the study also has a set of limitations. The test, for instance, was carried out with the help of an imaging equipment and a lot of the patients in the US group were already confirmed patients of breast cancer. Moreover, the AI tool should also improve patient care and needs regulatory approval.

SEE ALSO: We Now Have An AI Sex Robot That Can ‘Breathe’

Image for representation purpose only