Google shows how AI could detect lung cancer faster and more reliably

A new Google study shows how machine learning could one day be used to detect signs of lung cancer earlier than is often the case today.

Early warning: Danial Tse, a researcher at Google, has developed an algorithm that beats a number of qualified radiologists for testing. Tse and his colleagues formed a deep learning algorithm detect malignant lung nodules in more than 42,000 CT scans. The algorithms obtained revealed 11% fewer false positives and 5% fewer false negatives than their human counterparts. The work is described in a paper published in the journal Nature aujourd & # 39; hui.

Killer problem: Lung cancer has killed more than 160,000 people in the United States in 2018, making it the leading cause of cancer deaths. And while computed tomography (CT) can be an essential part of cancer screening, it is often not reliable either.

Great promise: Tse and colleagues argue that AI could help make lung cancer screening more reliable in the world, although they recognize that work needs to be validated on larger patient populations. Indeed, there is a growing interest in using AI to catch many types of cancer. Researchers have shown how machine learning can be used to detect both breast cancer and skin cancer, for example.

Small steps: These studies are exciting but need to be treated as small advances. It remains difficult to use AI in health care for confidentiality reasonsand because real-world data sets are rarely as perfect as those used in research studies.

It should also be noted that cancer treatment involves much more than simply detecting the disease. The determination of the right treatment, for example, may depend on many factors that vary considerably from patient to patient, making this part of the process much more difficult to automate.