Novel blood test uses AI to detect cancer before symptoms appear

Scientists have developed a blood test -- combining "liquid biopsy" with artificial intelligence -- that can detect cancer at its earliest stages.

The findings, published in the journal Nature, hold promise of being able to treat the disease long before symptoms appear, said Daniel De Carvalho, a senior scientist at University Health Network in Canada.

"A major problem in cancer is how to detect it early. It has been a 'needle in the haystack' problem of how to find that one-in-a-billion cancer-specific mutation in the blood, especially at earlier stages, where the amount of tumour DNA in the blood is minimal," said Carvalho.

By profiling epigenetic alterations instead of mutations, the team was able to identify thousands of modifications unique to each cancer type.

Then, using a big data approach, they applied machine learning to create classifiers able to identify the presence of cancer-derived DNA within blood samples and to determine what cancer type.

The scientists tracked the cancer origin and type by comparing 300 patient tumour samples from seven disease sites (lung, pancreatic, colorectal, breast, leukemia, bladder and kidney) and samples from healthy donors with the analysis of cell-free DNA circulating in the blood plasma.

In every sample, the "floating" plasma DNA matched the tumour DNA. The team has since expanded the research and has now profiled and successfully matched more than 700 tumour and blood samples from more cancer types.

The next steps to further validate this approach include analysing data from large population health research studies already under way in several countries, where blood samples were collected months to years before cancer diagnosis, researchers said.

Then the approach will need to be ultimately validated in prospective studies for cancer screening.