Google AI Spots Breast Cancer in Lymph Nodes 99% of Time

A Google artificial intelligence tool called LYNA can now spot if breast cancer has spread to lymph nodes with a 99 percent accuracy rate, the tech giant revealed last week.

When it comes to breast cancer treatment options, doctors base their decisions on whether the cancer has spread to the lymph nodes. It’s a vital, laborious determination with time constraints. Screenings may not be accurate and results may change.

Last year Google presented LYNA, which stands for Lymph Node Assistant, at the ISBI Camelyon Challenge held to encourage the evaluation of new and existing algorithms when it comes to detecting if cancer has spread to the lymph nodes.

According to the Camelyon Challenge’s website, “this task has a high clinical relevance but requires large amounts of reading time from pathologists.”

Google noted that while LYNA achieved a significantly higher cancer detection rate, an accurate algorithm alone was not enough to improve a pathologists’ workflow or the outcomes for breast cancer patients.

In two recently-published research papers – one in Archives of Pathology & Laboratory Medicine and the other in the American Journal of Surgical Pathology – Google said it developed a LYNA tool that assists pathologists in assessing such algorithms for adoption into their workflow and provides a 99 percent accuracy in detecting whether cancer had spread to the lymph nodes.

Based on the outcome, Google noted that AI such as LYNA could improve the accuracy of pathological diagnosis while reducing the “burden of repetitive identification tasks” and ultimately freeing up pathologists to focus on more challenging clinical tasks.

Scientists and technologists are increasingly turning their attention to AI and how it can be used to diagnose and treat cancer, as was the case at the prominent U.K. artificial intelligence conference, CogX, where healthcare was a major topic, Forbes reported.

Commenting on this, Dr Jack Kriendler of the Centre for Health and Human Performance, said he trusted AI more than his fellow doctors.

“I would sooner today trust computer scientists and data scientists to tell me how to treat a really complex system like cancer than my fellow oncologists,” he said. “I would not have said that two to three years ago. The regulatory bodies think very, very differently to that.”