New AI That Detects Breast Cancer Better Than Doctors

AI detecting breast cancer.
Google’s AI, the Lymph Node Assistant (LYNA), can detect breast cancer more accurately and swiftly than doctors. (Image: Screenshot via YouTube)

Futurists have been predicting for years that Artificial Intelligence (AI) will one day play a dominant role in healthcare. Google just gave us a glimpse of that future by revealing a new AI that detects breast cancer more accurately and swiftly than doctors.

Meet LYNA, a new AI that detects breast cancer

Pathologists usually detect breast cancer by inspecting the samples of lymph nodes taken from patients. The procedure reveals how aggressive the cancer is and the extent to which it has spread in the body. Google’s AI, the Lymph Node Assistant (LYNA), seeks to perform this task in a better way.

Created in association with the Naval Medical Center San Diego, LYNA is programmed to analyze scans of thousands of cancer patients and use that knowledge to identify signs of breast cancer in women. To train the AI, Google used cancer scans collected from medical centers in the Netherlands. The machine learning process enables the AI that detects breast cancer to achieve 99 percent accuracy when it comes to detecting cancerous cells.

Pathologists who used LYNA were found to perform better at detecting anomalies than those who didn’t use the AI tool. In fact, several pathologists claimed that LYNA enabled them to detect small cancerous growths easily, a task that would have taken a long time using traditional detection methods. “This represents a demonstration that people can work really well with AI algorithms than either one alone,” Business Insider quotes Yun Liu, a member of the team.

LYNA, a new AI that detects breast cancer better than doctors.
Pathologists who used LYNA were found to perform better at detecting anomalies than those who didn’t use the AI tool. (Image: Fernandozhiminaicela via Pixabay)

The research team now plans on testing the AI in a clinic where it would be fed with fresh scans of lymph nodes. If the AI proves its mettle and maintains its 99 percent accuracy rate with new data, LYNA could very well become a popular cancer detection tool in clinics worldwide in just a few years.

MIT’s AI technology for breast cancer

The Massachusetts Institute of Technology (MIT) is also developing AI that detects breast cancer. Around 335 high-risk lesions were tested on their AI, with the system accurately identifying 97 percent of breast cancers as malignant. This reduced the need for surgeries by over 30 percent as compared to existing diagnostics methods. MIT is collaborating with the Harvard Medical School and Massachusetts General Hospital (MGH) on the project.

“In the past, we might have recommended that all high-risk lesions be surgically excised. But now, if the model determines that the lesion has a very low chance of being cancerous in a specific patient, we can have a more informed discussion with our patient about her options. It may be reasonable for some patients to have their lesions followed with imaging rather than surgically excised,” an MIT article quotes Constance Lehman, chief of the Breast Imaging Division at MGH’s Department of Radiology.

The Massachusetts Institute of Technology (MIT) is also developing an Artificial Intelligence tool for cancer detection and diagnosis.
The Massachusetts Institute of Technology (MIT) is also developing an Artificial Intelligence tool for cancer detection and diagnosis. (Image: Screenshot via Youtube)

In addition to information on high-risk lesions, the AI looks at many different data points of the patient, like family history, pathology reports, demographics, and past biopsies. Since existing tools tend to be less accurate, doctors used to over-screen patients for breast cancer.

But with MIT’s machine-learning AI that detects breast cancer, doctors can prevent over-treatment and provide more targeted solutions for women. MGH is expected to start using AI in its clinic by next year. The team also plans to tweak the AI so that it can eventually be used for diagnosing other types of cancers.

Follow us on Twitter, Facebook, or Pinterest

RECOMMENDATIONS FOR YOU