Related Experiment Video
Updated: May 10, 2025
06:03
Integrating Augmented Reality Tools in Breast Cancer Related Lymphedema Prognostication and Diagnosis
Published on: February 6, 2020
6.4K
Implementing artificial intelligence in breast cancer screening: Women's preferences.
Alison Pearce1,2, Stacy Carter3, Helen Ml Frazer4,5
1The Daffodil Centre, The University of Sydney, A Joint Venture With Cancer Council New South Wales, Sydney, New South Wales, Australia.
Cancer
|April 22, 2025
Summary
Community preferences for artificial intelligence (AI) in breast cancer screening show a need for accuracy and faster results. Implementing AI against these preferences could decrease screening participation.
Area of Science:
- Medical imaging and diagnostics
- Health technology assessment
- Public health policy
Background:
- Artificial intelligence (AI) offers potential for enhanced breast cancer screening accuracy and efficiency.
- Public distrust in AI healthcare applications may hinder screening participation rates.
- Understanding community preferences is crucial for effective AI implementation in screening programs.
Purpose of the Study:
- To quantify community preferences regarding different models of AI implementation in breast cancer screening.
- To identify key factors influencing public acceptance of AI in healthcare settings.
- To assess the potential impact of AI implementation on screening participation.
Main Methods:
- An online discrete choice experiment survey was conducted with 802 eligible participants (aged 40-74) in Australia.
- Respondents evaluated screening options based on AI role, accuracy, ownership, representativeness, privacy, and waiting time.
- Analysis employed conditional and latent class models, willingness-to-pay, and predicted screening uptake.
Main Results:
- Participants favored AI that was more accurate, Australian-owned, representative, and offered shorter waiting times (all p < .001).
- Strong opposition was observed for AI used alone or for triage (p < .001).
- Acceptance of AI replacing a human reader required results 10 days faster; AI triage required 21 days faster.
Conclusions:
- Community preferences highlight the importance of accuracy, local ownership, and timely results for AI in breast cancer screening.
- AI implementation must align with public preferences to avoid significant reductions in screening participation (up to 22%).
- Addressing public concerns and preferences is vital for successful integration of AI into breast cancer screening pathways.
