Clinicians’ perspectives on the use of artificial intelligence to triage MRI brain scans

• DOI: 10.1016/j.ejrad.2025.111921
• Publication Date: 2025-01-06T17:17:21Z
• Authors: Munaib Din; Karan Daga; Jihad Saoud; David Wood; Patrick Kierkegaard; Peter Brex; Thomas C Booth
• Keywords: Triage

Artificial intelligence (AI) tools can triage radiology scans to streamline the patient pathway and also relieve clinician workload. Validated AI mitigate delays in reporting by flagging time-sensitive actionable findings. In this study, we aim investigate current stakeholder perspectives identify obstacles integrating clinical pathways. We created a survey ascertain of 133 clinicians across United Kingdom regarding acceptability an tool that triages MRI brain into 'normal' 'abnormal'. As part survey, supplied with information on training validation case numbers, model performance, using unseen data, explainability saliency maps. With regards specific use scans, 71% respondents preferred AI-assisted compared system without triage, typically chronologically. Notably, explained helped visualise model's decision making was found improve confidence. When shown heatmap, 60% participants felt more confident AI's decision. The results short communication demonstrate positive support for implementation AI-assistive triage.