Work

Impact of Exposure Parameters on Deep Learning Models in Chest Radiography and Implications for Deployment

Year: 2026

Type: article

Abstract: An algorithmic framework was developed to identify and quantify shortcut learning and bias driven by exposure parameters in chest radiographs, revealing hidden sources of bias in medical artificial intelligence.

Source: Radiology Artificial Intelligence

DOI: 10.1148/ryai.250731

Authors: Han-Jay Shu, Shun-Ting Chang, Rodrigo Rosa Gameiro, Judy Wawira Gichoya, Leo Anthony Celi, Po-Chih Kuo

Institutions: National Tsing Hua University, Beth Israel Deaconess Medical Center, Massachusetts Institute of Technology, Emory University, Harvard University

Language: English

Topic: Artificial Intelligence in Healthcare and Education

Subfield: Health Informatics

Field: Medicine

Domain: Health Sciences

SDG: Good health and well-being

Open Access status: green

Funders: National Science Foundation, National Science and Technology Council, National Institutes of Health, National Heart, Lung, and Blood Institute, Fogarty International Center

Awards: A Learning Ecosystem for Teaching High School Students Machine Learning Concepts and Data Science Skills in Healthcare and Medicine, Bridge2AI: Patient-Focused Collaborative Hospital Repository Uniting Standards (CHoRUS) for Equitable AI, G5217900148, Developing a Hive Learning and Datathon Supported Course on Imaging and Multimodal Data for Resource-Limited Institutions (CIMDAR-HIVE), G782677697, G816157098, MUST Data Science Research Hub (MUDSReH), Overcoming inequities in Pulse oximetry Through clinical InformatiCs (OPTIC)

OpenAlex ID: W7165805441

FWCI: 1.861

Cites: 18

Cited by: 1

Related to: