Underdiagnosis bias of artificial intelligence algorithms applied to chest radiographs in under-served patient populations

Artificial intelligence (AI) systems have increasingly achieved expert-level performance in medical imaging applications. We find that classifiers produced using advanced computer vision techniques consistently and selectively underdiagnosed under-served patient populations and that the underdiagnosis rate was higher for intersectional under-served subpopulations, for example, Hispanic female patients.

⚡ This is an original paraphrased summary — not copied from the abstract. Full paper available at the source link below.

• 1 However, there is growing concern that such AI systems may reflect and amplify human bias, and reduce the quality of their performance in historically under-served populations such as female patients, Black patients, or patients of low socioeconomic status.
• 2 Such biases are especially troubling in the context of underdiagnosis, whereby the AI algorithm would inaccurately label an individual with a disease as healthy, potentially delaying access to care.
• 3 Here, we examine algorithmic underdiagnosis in chest X-ray pathology classification across three large chest X-ray datasets, as well as one multi-source dataset.

This research advances how AI systems learn, reason, and solve problems — with direct implications for automation and scientific discovery.

This summary is based on publicly available metadata and abstract. For the full research paper, visit the original source: