Google's AI Can Now Detect Breast Cancer Better Than Humans Do

Joanna Estrada
October 17, 2018

This is one of the reasons why almost half a million women die worldwide and are caused by breast cancer, which is estimated to be 90% as the result of metastasis. However, researchers at San Diego Medical Center and Google AI (Artificial Intelligence Research Division) have developed a promising solution using cancer detection algorithms and self-assessing lymph node biopsies. Finding them is a time-intensive and hard task for pathologists.

Despite misidentifying giant cells, germinal cancers and histiocytes, LYNA managed to perform better than a pathologist.

Even more impressively, the machine was able to accurately pinpoint cancers and other "suspicious regions" within each slide-some of which were too small for humans to catch. The results were remarkable when LYNA worked as a companion to pathologists.

The scientific report published on this issue explains that: "Artificial Intelligence Algorithms can exhaustively evaluate each occurrence, individually and in detail".

On top of this incredible accuracy, a separate study conducted by Google found that its AI technology can help pathologists cut the amount of time it takes them to review slide samples by half.

To ideal this Google AI tool, the search giant used a de-identified data set of breast cancer patients' lymph node scans from medical centers in the Netherlands.

Google's AI tool LYNA is based on Inception-v3, an open-source image-recognition deep-learning model.

The team behind the test was able to train its algorithm named LYmph Node Assistant or LYNA to recognise the characteristics of tumours from two specific slides. It also has yet to be used in real-life clinical situations. If and when it's ready for practical use, though, it could both lead to extremely reliable diagnoses and free doctors to focus more of their time on caring for their patients. Additionally, Verily, which is Alphabet's life sciences subsidiary, is developing a system that determines a person's risk of heart disease using retinal scans.

Other reports by Click Lancashire

Discuss This Article