Publications by authors named "Qinghan Xue"
EClinicalMedicine
April 2024
Article Synopsis
- Artificial intelligence in healthcare often reflects existing historical inequities, prompting the need for a new framework to evaluate fairness in AI performance for different patient populations.
- The Health Equity Assessment of machine Learning performance (HEAL) framework was developed to quantitatively analyze whether health AI tools better serve those experiencing worse health outcomes through a detailed four-step method.
- In a case study involving a dermatology AI model using diverse teledermatology cases, the HEAL metric was used to assess the likelihood that the AI performed better for groups with poorer health outcomes, indicating its potential for promoting equity in AI technologies.
