Frameworks to assess and mitigate bias in artificial intelligence
As the adoption of artificial intelligence (AI) increases in healthcare, our team believes that it will be important to understand whether these tools perform equitably for people of all backgrounds. Our team is exploring appropriate metrics for assessing the equitability of AI model performance. The goal is to develop methods for mitigating any detected bias. The team also seeks to apply these insights to AI models that are developed and deployed.
Collaborating faculty