Patient and Public Perceptions of Artificial Intelligence in Breast Imaging and Clinical Decision-Making: An Exploratory Cross-Sectional Survey Study

Abstract

Background/Objectives: Artificial intelligence (AI) shows promise in supporting mammography interpretation and triaging referrals, potentially enhancing breast screening. However, successful AI integration depends on patient acceptance and trust. This study explores patient and public perceptions of AI in breast imaging and clinical decision-making to identify knowledge gaps and guide communication strategies. Methods: Paper surveys were distributed to women attending the Breast Care Unit at Queen’s Hospital, Burton, and the London Breast Institute between August and December 2025. Demographic data, levels of trust and comfort with AI, and concerns about AI were collected. Responses were analysed using descriptive statistics, Pearson’s Chi-square tests with Cramér’s V and thematic analysis. Results: One hundred and twenty participants completed the survey. Fifty percent would accept AI alongside clinicians for interpretation of mammograms or ultrasound scans, significantly associated with no previous breast cancer diagnosis (p = 0.02; Cramér’s V = 0.22, 2 degrees of freedom (df)) and technological comfort (p < 0.001; Cramér’s V = 0.42, 1 df). Lower acceptance was found among those with prior diagnosis and low comfort with technology. Acceptance of AI-assisted triage (44.5%) was also significantly associated with technological comfort (p = 0.008; Cramér’s V = 0.30, 1 df). Eighty percent reported no knowledge of AI use in breast clinics, and only 37% would trust AI findings. Qualitative analysis identified three themes: (1) clinician oversight as indispensable, (2) the knowledge gap as a barrier to acceptance, and (3) concerns about operational risks and accountability. Conclusions: Although patients were generally receptive to AI, acceptance was conditional on clinician supervision. Limited awareness and concerns about diagnostic accuracy remain barriers to implementation. Educational initiatives should precede widespread adoption to support informed and confident patient acceptance of AI-assisted imaging and decision-making.

Ähnliche Arbeiten