RELATED POST
Google Health outlined its mission in November to “help everybody live their healthiest life.” Today, the division released “initial findings” on using AI to improve breast cancer screening.
Google notes how “spotting and diagnosing breast cancer early remains a challenge.” Currently, digital mammography is used to detect it, but reading breastx-ray images is a “difficult task, even for experts, and can often result in both false positives and false negatives.”
In turn, these inaccuracies can lead to delays in detection and treatment, unnecessary stress for patients, and a higher workload for radiologists who are already in short supply.
The approach of the company includes the use of artificial intelligence. Findings made over the past two years have been published in Nature today:
These findings show that our AI model spotted breast cancer in de-identified screening mammograms (where identifiable information has been removed) with greater accuracy, fewer false positives, and fewer false negatives than experts.
This follows work on detecting metastatic breast cancer from specimens of the lymph node in 2017, and deep learning algorithms that help doctors diagnose breast cancer.
Google Health partnered with DeepMind Alphabet Group, Cancer Research UK Imperial Center Northwestern University, and Royal Surrey County Hospital to “see if artificial intelligence could support radiologists to spot the signs of breast cancer more accurately.”
In this evaluation, our system produced a 5.7% reduction of false positives in the US, and a 1.2% reduction in the UK. It produced a 9.4% reduction in false negatives in the US, and a 2.7% reduction in the UK.
What is surprising is how the AI system, which doctors would normally use, did not have access to patient history and past mammograms. The model was trained from 76,000 women in the UK and 15,000 women in the US de-identified mammograms.
In an independent study of six radiologists, the AI system outperformed all of the human readers: the area under the receiver operating characteristic curve (AUC-ROC) for the AI system was greater than the AUC-ROC for the average radiologist by an absolute margin of 11.5%.
Next is further testing, as well as “prospective clinical studies and regulatory approval” on how AI could help detect breast cancer. Google hopes that “machine learning research will be translated into tools that benefit clinicians and patients” in the “coming years.”