0822 GMT July 21, 2019
Researchers from the University of North Carolina used artificial intelligence machine learning to single out and classify tumors based on grade, estrogen receptor status, PAM50 intrinsic subtype, histologic subtype and risk of recurrence score, UPI reported.
In a head-to-head comparison with two pathologists' findings, the AI-based technology differentiated low-intermediate and high-grade tumors with 82 percent accuracy. The pathologists were correct with 89 percent of their tumor grade assessments.
"Your smartphone can interpret your speech, and find and identify faces in a photo," Heather D. Couture, a graduate research assistant in the UNC-Chapel Hill Department of Computer Science and the study's first author, said in a press release. "We're using similar technology where we capture abstract properties in images, but we're applying it to a totally different problem."
The AI technology could also tell with high accuracy whether ductal and lobular tumors it found had a high or low risk for recurrence. It also spotted the basal-like subtype that determines gene expression within a breast cancer tumor.
"We were surprised that the computer was able to get a pretty high accuracy in estimating biomarker risk just from looking at the pictures," said UNC Lineberger's Melissa Troester, a professor in the UNC Gillings School of Global Public Health. "We spend thousands of dollars measuring those biomarkers using molecular tools, and this new method can take the image and get 80 percent accuracy or better at estimating the tumor phenotype or subtype. That was pretty amazing."
The researchers considered how the computer is able to identify these tumors and their various features, and if it can forecast future results.
"The computer extracted a lot of information from the images," Troester said. "We would like to test how well these features predict outcomes, and if we can use these features together with things like molecular data to do even better at giving patients a precise view of what their disease course looks like, and what treatments might be effective."
"This has a long way to go in terms of validation, but I think the accuracy is only going to get better as we acquire more images to train the computer with," Charles M. Perou, distinguished professor of Molecular Oncology at the UNC School of Medicine, said.
The findings come from a study published in September in the journal npc Breast Cancer.