Related Experiment Videos
CLEAR-AI: confounder-aware learning for equitable and accurate reasoning in AI for diagnosis
Vedant Joshi1, Ramon Correa1, Avisha Das2
1Arizona State University, School of Computing and Augmented Intelligence, Tempe, Arizona, United States.
Purpose:
A critical challenge impeding the deployment of artificial intelligence (AI) models in healthcare lies in implicit bias against multiple correlated sensitive attributes.
Approach:
We developed a multibranch adversarial debiasing approach that can debias a multilabel diagnostic model simultaneously for correlating confounding factors using a dynamic weighted gradient reversal technique to reduce disparities. The proposed methodology was evaluated on chest X-ray imaging data-trained on CheXpert and evaluated on two external datasets [MIMIC-CXR (Beth Israel, Massachusetts, United States) and Emory Healthcare (Atlanta, Georgia, United States)].
Results:
The disease classification performance of the proposed debiased method significantly overlaps with the baselines, indicating no drop in the task performance. We found the debiased model to reduce TPR and FPR disparities across multifactor subgroups [age, race, and support device(s)] while maintaining overall task performance.
Conclusions:
Although we focused on three specific confounders, the proposed adversarial debiasing framework readily extends to account for an arbitrary number of sensitive variables. The findings highlight the promising potential of adversarial training techniques to enhance fairness and trustworthiness in the deployment of AI models in diverse healthcare settings.
