Artificial Intelligence in Radiology: Perceptions, Adoption Barriers, and Trust Among Iranian Radiologists in a Global Context

Artificial intelligence (AI) is transforming radiology globally, yet adoption varies significantly across regions due to cultural, educational, and infrastructural factors. This study examines Iranian radiologists’ perceptions, trust, and barriers to AI adoption through a cross-sectional survey of 128 professionals (radiologists, residents, and technologists) from diverse healthcare settings. Results revealed cautious optimism: 78.1% anticipated AI would significantly impact radiology within a decade, primarily as a workflow optimizer (69.5%) or second reader (73.4%). However, critical barriers emerged, including lack of formal AI training (77.3% had none), low confidence in AI tools (mean score: 2.35/5), and concerns about reliability (52.3%) and legal accountability (46.1%). Only 29.7% trusted AI-generated reports (90% accuracy), with 83.6% demanding mandatory human oversight. Demographic differences were notable; younger professionals (<35 years) were more optimistic about AI’s augmentative role (p < 0.05). These findings align with trends in low- and middle-income countries (LMICs), where limited training and infrastructure hinder adoption compared to high-income regions. The study highlights urgent needs: integrating AI into radiology curricula, pilot programs to build trust, and regulatory frameworks addressing transparency and liability. By addressing these challenges, Iran could leverage AI’s potential while navigating LMIC-specific constraints. This research contributes to global discourse on equitable AI adoption by contextualizing Iran’s position alongside international benchmarks.