To gain insights into patients’ perspectives on risks and benefits of AI in radiological diagnostics, conditions for acceptance, trust in AI diagnoses, expectations from physicians, and the underlying motives for their attitudes towards AI.
Key Findings:
AI models can classify medical images with high accuracy, often matching or exceeding radiologists' performance, which has significant implications for patient care.
Patients express skepticism about AI in healthcare, fearing misdiagnosis and questioning the quality of AI training data.
Patients prefer AI to support rather than replace physicians in diagnostic processes.
Transparency and communication about AI use are crucial for patient acceptance and trust.
Interpretation:
Patients' acceptance of AI in radiology is influenced by their trust in the technology and the physician-patient relationship, with a preference for AI as a supportive tool rather than a standalone system, highlighting the need for effective communication.
Limitations:
The study primarily reflects the views of participants from high-income countries, which may not be generalizable to other contexts, particularly in terms of cultural differences.
Limited understanding of the specific characteristics of the physician-patient relationship that foster trust in AI.
Conclusion:
The study highlights the importance of addressing patient concerns and expectations regarding AI in radiological diagnostics to enhance acceptance and trust.
