An international team of scientists may have made a major breakthrough in early cancer detection: a blood test that finds more than 50 types of cancer, including specific locations in the body.

They did it with a little help from artificial intelligence.

The research — published this week in the Annals of Oncology — involved administering a blood test to about 15,000 participants and zeroing in on the DNA of cancer.

When tumor cells die, they are flushed into the bloodstream. Knowing this, the researchers designed the test to look specifically for the DNA of those dead cells. Chemical changes to the structure of that DNA revealed — to a high degree of accuracy — the type and location of cancer.

What is DNA methylation and why does it matter?

melanoma photomicrograph, cancer detection blood test and methylation patterns The new test could make a big difference in lead time for patients. The algorithm could discern between methylation changes that are cancerous or non-cancerous, and it could even pinpoint the tissue of origin before the onset of symptoms. (Photo: BonD80/Shutterstock)

To help them decipher those DNA chemical changes, called methylation patterns, researchers developed a machine learning program. Essentially, the algorithm sifted through those methylation changes, learning to distinguish between cancerous and non-cancerous — and then identifying them appropriately.

"The test exploits the fact that cancers are abnormally methylated, and ... that different cancers have different methylation patterns," study co-author Eric Klein from the Cleveland Clinic in Ohio, tells Chemistry World.

Overall, the test accurately predicted the location where the cancer began more than 90% of the time, and only wrongly indicated the presence of cancer 0.7 percent of the time.

Currently in clinical trials, the test could be the only option for early detection when it comes to particularly dangerous forms of the disease. It also means some types of cancer can be spotted before symptoms appear.

"You need to use a test like this in an independent group at risk of cancer to actually show that you can find the cancers, and figure out what to do about it when you find them," Geoffrey Oxnard of Boston's Dana-Farber Cancer Institute, part of Harvard Medical School, tells The Guardian.

Initial results for cancer detection are encouraging

While there have been several promising developments in early detection recently — tests based on a patient's breath or using crumpled graphene — no examination has been able to sniff out so many forms of cancer at once.

And when it comes to fighting the disease, early detection is crucial.

"Cancer is a complicated disease, [and currently] every type has a different testing and screening system," Abu Siuna, a cancer researchers who was not involved with this study, told CNN in December. "In most cases, there is no general test to test their status."

That may be about to take a dramatic change if this new blood test continues to impress in clinical trials.

"This is probably the best data we have so far — I haven't seen anything better than this," David Polsky, a New York University researcher who wasn't involved in this study tells Chemistry World. "But it's still not really ready to be deployed into the general population without more testing."

This new blood test accurately predicts 50 types of cancer
Scientists using artificial intelligence have devised a highly accurate blood test for 50 kinds of cancer by looking for the DNA of dead tumor cells.