IBM's artificial intelligence platform Watson is being trained to help cancer patients receive more personalized care by identifying the specific genetic mutations causing their disease. With Watson's reports, healthcare providers will be able to develop and administer therapies to directly treat each patient.
In a current example, pathologists studied melanoma biopsies for a single BRAF mutation, which is shown in 50 percent of melanomas. While there are drugs that target the mutation, the patient must undergo broad-spectrum chemotherapy if the BRAF is not identified. Watson is able to quickly identify the different mutations of BRAF as well as a dozen other genes that lead to skin cancer in order to provide a complete individual report on each patient.
"We have to change our whole behavior in looking at tumors. We are missing too much and too often treatment does not work for patients," said Ravindra Kolhe, MD, PhD, director of the Georgia Esoteric & Molecular Labs in the Department of Pathology at the Medical College of Georgia at Augusta University. "With Watson, all oncogenes become targets. The majority of the time, we just tell patients they have a cancer, Watson can help us provide more comprehensive, personalized care to patients."
In tests on Watson's ability to provide comprehensive and accurate patient reports, the AI was able to provide a 30-page assessment on a patient with a rare soft tissue sarcoma. The report included information on seven different gene mutations, where the mutations occurred and the best treatment options. Overall, Watson was able to deliver a report that would have taken 10 people 10 days.
"In about 20 minutes, Watson looks at what is abnormal in the sample, then takes the abnormalities and looks all around the world to see what are the drugs already used against them, what studies are underway against them, even other drugs out there that might be useful," said Kolhe.