UCLA Researchers Develop AI Tool to Improve Alzheimer's Detection Among Underserved
UCLA Health
|
Contributed by: Kate Gamble
Summary
Researchers at UCLA have created an innovative AI tool to uncover undiagnosed cases of Alzheimer's disease, focusing particularly on underrepresented communities prone to underdiagnosis. Published in npj Digital Medicine, the study shows that African Americans and Hispanic/Latino individuals are frequently diagnosed later than non-Hispanic whites, highlighting existing disparities in Alzheimer’s care. The new model, which employs a semi-supervised positive unlabeled learning approach, achieved sensitivity rates between 77% and 81% across diverse ethnic groups, significantly surpassing traditional methods. This development is crucial for healthcare professionals, as it not only enhances diagnostic accuracy but also aims to reduce racial and ethnic disparities in disease recognition and treatment.
