Google says it has succeeded in improving breast cancer detection with the help of artificial intelligence. The tech company published its findings after a two-year study on breast cancer screening in the scientific journal Nature.
For the past two years, Google has been working closely with leading scientists in the US and the UK to see how artificial intelligence can improve early detection of breast cancer. By enabling an AI model to analyze mammograms, breast cancer could be detected with greater accuracy, according to Google. There were fewer false positives and fewer false negatives than the mammograms analyzed by specialists.
More than 42 million mammograms are made annually in the Americas and the UK alone. In the UK, more than 55,000 women are diagnosed with breast cancer every year. In the US, one in eight women will experience the disease in her lifetime. Even for specialists it remains a complicated task to analyze the images, which often leads to a wrong diagnosis. These mistakes can lead to treatment delays, unnecessary patient distress and increased workload for radiologists. A faster and more accurate analysis of the X-rays is therefore of great importance, according to Google.
During the two-year study, anonymized mammograms of nearly 120,000 women were analyzed. In the US, the number of false positives decreased by 5.7 percent thanks to the AI model, in the UK by 1.2 percent. The number of false negatives fell by 9.4 percent in the US. In the United Kingdom, a decrease of 2.7 percent was noted.
As a result of the positive initial research results, Google will further develop the AI model. New applications may be able to assist radiologists in the future with breast cancer screening itself.
Microsoft also uses artificial intelligence to detect cancer at an early stage. The company announced in November last year that it had developed an AI tool together with the Indian SRL Diagnostics to detect cervical cancer more quickly.