Evidence for AI models possibly having a racial medical bias was found in a recent University of Michigan study. Researchers have found that AI models may misdiagnose patients of color.
Associate Professor of Computer Science and Engineering Jenna Wiens says this is due to patients of different racial and ethnic backgrounds receiving medical tests that others are not. She says testing disparity results in underrepresentation of certain racial groups in diagnostic data.
“If these data contain spurious correlations or bias, the models trained on these data can replicate or even amplify these biases. So, it’s important to scrutinize the quality of the data.”
Wiens says she along with her fellow co-authors propose human reviews of AI diagnoses and closing racial disparities in medical testing.
Non-commercial, fact based reporting is made possible by your financial support. Make your donation to WEMU today to keep your community NPR station thriving.
Like 89.1 WEMU on Facebook and follow us on Twitter
Contact WEMU News at 734.487.3363 or email us at studio@wemu.org
