Google AI system found to identify breast cancer more accurately than radiologists

Breast cancer is the most common cancer in the UK – about one in seven women in the UK develops breast cancer during their lifetime, contributing to around 55,000 diagnoses every year, according to Cancer Research UK. There could soon be a tool to “support” radiologists performing breast cancer screenings, as an artificial intelligence (AI) system developed by Google Health, in collaboration with the charity, was found to better detect breast cancer from mammograms. A widely used breast cancer screening tool, screening mammography involves X-ray images taken of the breast from multiple views and analysed by at least two radiologists in the UK or one in the US.

In a recent study, the AI model processed mammograms – with no extra patient information – from a selection of more than 25,000 women in the UK and over 3,000 women in the US for signs of breast cancer. This resulted in a 1.2% reduction in false positives and 2.7% reduction of false negatives from the AI model compared to radiologists in the UK; there was a 5.7% reduction in false positives and 9.4% reduction of false negatives in the US. A false positive result is described when a mammogram is reported as abnormal when no cancer is present while a false negative is when a mammogram is reported as normal even though breast cancer is present.

“These research findings suggest that we are on our way to developing a tool that can help clinicians spot breast cancer with greater accuracy,” said a proud Dominic King, UK Lead at Google Health, although he admitted that further testing, clinical validation and regulatory approvals were still required. Similarly, Michelle Mitchell, Cancer Research UK’s Chief Executive, thinks that AI could be the future of cancer diagnosis and care. She said “embracing technology like this” could significantly improve screening diagnoses, when treatment is more likely to be successful and ensures more people survive the disease.

Category: Features, Technology & Devices