False negatives to false positives have a definitive ratio, but they used to be accurate, at least in the field of medicine… till Google entered the game, in all mightiness came in as the Almighty. Google is developing artificial intelligence to help doctors identify breast cancer; ergo, it has hit ground with this homegrown AI model.
The project is a fruit of transatlantic research collaboration with Google Health, where they worked jointly with the Cancer Research UK Imperial Center, Northwestern University, Royal Surrey County Hospital, and DeepMind, Google’s parent Alphabet Inc.’s AI development unit.
Reading mammograms have always been a difficult task, even for experts, and certain inaccuracies can lead to the delay of detection and subsequent treatment, which leads to unnecessary stress for patients and increase workloads for radiologists who are already fighting short supply.
Google’s AI model spotted breast cancer in de-identified mammograms (where identifiable information has been removed) with fewer false positives and fewer false negatives and with utmost accuracy.
X-ray imaging for Breast Cancer or mammogram to date remains the most common method to screen for breast cancer, with over 42 million exams performed each year in the UK and the US combined.
According to a blog post published by Google, at first, the team used AI to scan X-ray images, then looked for signs of early breast cancer by identifying changes in the breast for over 28,000 women. Then they went on to compare the computer’s guesses against the actual medical outcomes.
And guess what happened? The systems weren’t perfect initially, but ultimately, the experts were able to reduce false negatives by 9.4% and cut down false positives by 5.7% for women in the US. In the UK, where 2 radiologists double-checked the results, the model cut down false negatives by 2.7% and reduced false positives by 1.2%.
So, moral of the story? The machine outperforms doctors in identifying breast cancer.
And the question, can it replace doctors? The answer: NOPE!
In some instances, doctors flagged cancer where the model initially missed. Google has been cautious to frame this project with intrinsic help from radiologists, not replacing them. According to the experts, they are both co-dependent as there have been cases where the radiologists caught something that the machine missed and vice versa. Bringing the two together could strengthen the overall result, and the ultimate goal is to defeat cancer and fight for a cancer-free society.
The project is an ongoing part of Google to expand into the field of healthcare, and the tech giant partnered with Ascension to gain access to health records of millions of American citizens. This may be a revolutionary step toward defeating cancer and choosing life over the terminal illness!