Related Experiment Video
Updated: Jan 11, 2026
05:33
Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System
Published on: July 11, 2025
773
Radiologists' Perspectives on AI Integration in Mammographic Breast Cancer Screening: A Mixed Methods Study.
Serene Si Ning Goh1,2,3, Qin Xiang Ng2, Felicia Jia Hui Chan2
1Department of General Surgery, National University Hospital Singapore, Singapore 119077, Singapore.
Cancers
|November 13, 2025
Summary
Radiologists in Singapore view artificial intelligence (AI) as a helpful tool for breast cancer screening, but not a replacement for human expertise. Successful AI integration requires clear guidelines, local data validation, and user training.
Area of Science:
- Medical Imaging
- Artificial Intelligence in Healthcare
- Radiology
Background:
- Artificial intelligence (AI) shows promise for enhancing breast cancer screening accuracy and efficiency.
- Real-world AI adoption in national screening programs is limited, especially regarding Asian radiologists' views.
- Understanding radiologists' perspectives is crucial for effective AI integration in Singapore's breast screening program.
Purpose of the Study:
- To explore radiologists' perceptions of AI adoption in Singapore's breast screening program.
- To identify perceived benefits, barriers, and requirements for safe AI integration.
- To assess radiologists' attitudes towards AI's diagnostic role and integration preferences.
Main Methods:
- Mixed methods study combining a cross-sectional survey (n=17) and semi-structured interviews (n=10) with experienced radiologists.
- Survey measured confidence, attitudes, and integration preferences regarding AI-assisted mammography.
- Thematic analysis of interviews guided by the Unified Theory of Acceptance and Use of Technology (UTAUT).
Main Results:
- Most radiologists (64.7%) supported AI as a companion reader, but only 29.4% found its performance comparable to humans.
- Radiologist confidence in AI was highest when validated on local datasets.
- Key concerns included false positives, workflow issues, medico-legal accountability, and costs; trust hinges on national guidelines, local validation, and defined roles.
Conclusions:
- Radiologists advocate for AI as an adjunct, not a replacement, in breast cancer screening.
- Successful AI adoption necessitates robust regulatory frameworks and seamless workflow integration.
- Transparent local data validation and structured user training are essential for safe and effective AI implementation.
