Google-backed study shows artificial intelligence can be used to detect lung cancer

Machine learning software helps computers analyse data and detect patterns

Google researchers teamed up with medical centres in the US to see if AI could be used to detect lung cancers. Reuters

A study by Google and several medical centres in the US shows artificial intelligence (AI) can be used to detect lung cancer and provide doctors with more accurate readings of computerised tomography (CT) scans.

"While the vast majority of patients remain unscreened, we show the potential for deep learning models to increase the accuracy, consistency and adoption of lung cancer screening worldwide," said a report of the study published in the Nature journal.

Although the technology is still under development and not ready to be implemented, the research suggests computers can be trained to recognise patterns linked to a specific medical condition and detect it in CT scans.

In the Google study, researchers applied AI to CT scans used to screen people for lung cancer, which caused around 1.7 million deaths worldwide last year and is projected to cause 142,000 deaths in the US in 2019, according to the American Cancer Society.

CT scans are recommended for people at high risk due to a long history of smoking. However, tests are not always accurate and can miss tumours or mistake benign tumours to be malignant, leading to  unnecessary invasive procedures or surgery.

While different radiologists may have their own interpretations of an individual scan, the Google study shows computers were as good as, or better than, doctors at detecting small lung cancer on CT scans.

AI involves using accumulated data and software known as "machine learning" to enable computers to mimic human thought processes. The more data is fed into the computer, the easier it is to recognise patterns. Machine learning – or deep leaning – means the computer essentially becomes smarter over time as it stores and organises data and applies contextual intelligence to the process.

Google has been actively developing machine-learning software, and in 2014, acquired a start-up called DeepMind for a reported $400 million (Dh1.46bn) to intensify its research in this field.

The tech company is also working on numerous applications of AI in medicines, including the detection of diseases and streamlining health insurance coverage. Google has created systems to help pathologists read slides to diagnose cancer and assist ophthalmologists to detect eye disease.

"We have some of the biggest computers in the world," Daniel Tse, a project manager at Google, told The New York Times in an article published on Monday. "We started wanting to push the boundaries of basic science to find interesting and cool applications to work on."