As Digital Journal’s Ken Hanly reported, Google has developed artificial intelligence aimed at helping doctors identify breast cancer more accurately and faster. The algorithm scans mammograms reduces false negatives by 9.4 percent (for U.S. patients) and 2.7 percent (for U.K. patients), according to experimental data.
Then new technology comes from DeepMind and Google Health. The artificial intelligence is not only designed to reduce error rates, it also has the intention of enabling doctors detect breast cancer early. This is based on six radiologists assessing 500 randomly-selected mammograms.
According to expert Joseph Mossel, CEO and co-founder of Ibex Medical Analytics: “The Deep Mind algorithm utilities the same technology that we use – deep learning – to improve the accuracy and efficiency of diagnoses at a different stage of the cancer pathway.”
He notes that while “Google’s technology is chiefly concerned with radiology which is the initial detection of whether cancer is present or not£, his company’s own “Second Read System looks at pathology which is the final diagnosis of what is suspected and any subsequent grading and is already used in a live clinical setting in several labs globally.”
Commenting on the potential benefits of artificial intelligence for healthcare, Mossel says: “Both systems address the same underlying problem – a chronic shortage of trained staff which is putting an increased strain on the current system and is impacting the quality of diagnoses.”
Taking the U.K. as an example, Mossel notes: “Just 3 percent of NHS labs have enough pathologists, which means that 97 percent don’t have sufficient resources. This can lead to error rates – where benign diagnoses are actually cancerous – as high as 12 percent; a truly alarming statistic.”
Extolling the overall benefits, Mossel concludes: “With more than two million new cases identified each year and as many as one in eight women diagnosed with breast cancer, it is a disease that impacts many lives. But with more people leaving the professions that help identify cancer than joining, it is vital that we empower those left with the tools to do their jobs as effectively as possible and AI technology can do just that.”
