Google Develops AI Tool That Detects Breast Cancer

It was only a matter of time until Google dipped its toes in medical innovation by working on advanced machine-learning apps for medical breakthroughs.

LYNA AI for Cancer Diagnosis

Google has seemingly achieved the impossible by developing the most up-to-date AI algorithm called the LYNA, or Lymph Node Assistant. The revolutionary LYNA AI is almost 99% accurate at being able to detect metastatic breast cancer. What’s more, it can even pinpoint the exact source of the cancerous cell formations and other “suspicious regions within Lymph nodes” in the human body. Providing patients with a greater chance at recovering.

“Artificial intelligence algorithms can exhaustively evaluate every tissue patch on a slide,” the paper concludes. “We provide a framework to aid practicing pathologists in assessing such algorithms for adoption into their workflow (akin to how a pathologist assesses immunohistochemistry results).”

According to Google’s research, a study published in the JAMA Journal of Medicine, cites that in around 62% of the cases, doctors when unaided by technology, and under the pressure of the race against time often fail to make a correct diagnose of metastatic cancer in patients.

Google conducted a research where six board-certified pathologists used LYNA in their medical practice, the study result yielded that the AI not only cut down the review time needed for each case, but also improved their diagnosis and eliminated the chance of a misdiagnosis.

“LYNA achieves higher tumor-level sensitivity than, and comparable slide- level performance to, pathologists,” the researchers stated. “These techniques may improve the pathologist’s productivity and reduce the number of false negatives associated with morphologic detection of tumor cells.”

Human-in-the-Loop Automation Concept

Gene Munster, LoupVentures founder and former Piper Jaffray analyst correctly surmises that Google’s new innovation will enable doctors to “more accurately diagnose conditions in less time” via the unique combination of human and machine.

He further elaborated that the AI invention was a part of a much bigger and burgeoning concept known as Human-in-the-loop Automation, where doctors and health professionals come together to work along with innovative, machine-learning technologies, like LYNA, to enhance their capabilities and reduce human error.

“Human-in-the-loop automation involves human labor augmented – not replaced – by machines. In this case, a machine would make a first pass. Screening samples and flagging possible positives for human review.” Munster said.

About the details of the way the LYNA works, VentureBeat reports:

“LYNA is based on Inception-v3, an open source image recognition deep learning model that’s been shown to achieve greater than 78.1 percent accuracy on Stanford’s ImageNet dataset. It takes as input a 299-pixel image (Inception-v3’s default input size), outlines tumors at the pixel level, and, in the course of training, extracts labels — i.e., predictions — of the tissue patch (“benign” or “tumor”) and adjusts the model’s algorithmic weights to reduce error.”

Future of LYNA

It’s a long road ahead for an AI like LYNA considering it will have to jump over various hurdles of clinical trials and FDA approvals before being available for the masses.

However early results of the studies conducted show great promise for the AI despite certain restrictions. One can hope for a future where cancer can be yanked out before it takes root in the body with the help of LYNA.