How is artificial intelligence being used to trace the origin of cancer?

A Curie Institute has developed an artificial intelligence capable of identifying cancers that cannot be detected by conventional means.

The use of artificial intelligence (AI) in medicine is beginning to emerge with cutting-edge solutions. In the case of Wilfred, a 50-year-old patient, it was AI that discovered the origin of the cancer despite the doctors’ examinations.

Wilfrid lives in Cherbourg. In 2019, he was teaching a class when he suddenly had a seizure. When he woke up in the hospital, he learned that he was suffering from cancer and had two brain tumors.

“The analyzes could not, by human eye, determine the primitive origin of my cancer,” Wilfred testifies.

He underwent surgery successfully but unfortunately tumors appeared in the brain as well as in the back. In view of how widespread the tumors were, the medical profession decided to begin chemotherapy. Then Wilfred integrates the CAPI protocol: cancer with unknown primary history, which uses artificial intelligence.

Deal with the tool as quickly as possible

In this way, the AI ​​was able to identify what doctors could not say for sure. In December 2022, the AI ​​has 90% concluded that the kidneys are the cause of the cancer. The software generates hypotheses from a database in which more than 20,000 RNA profiles of various tumors have been entered.

“We analyzed and sequenced these tumors and sent the profiles to our AI tool. In 80 percent of the cases, the tool was able to give us a diagnosis of the original tissue with very high accuracy scores,” Sarah Watson, MD, a Curie Institute physician, explains to BFMTV.

For her, the medical profession must fit into the instrument as quickly as possible: “We must be able to use this instrument as we have learned to use a stethoscope, and radiology devices.”

She also asserts that it will be essential to be able to identify the limits and biases of AI machines. At present, 80 people have integrated the CAPI protocol.