Skip to content

CT interpretation wait times with artificial intelligence triage across emergency, inpatient, and outpatient settings

Oct 2026 · NanoScience + Engineering · Vol 14198, pp. 141980F - 141980F-5 · 0 citations · 6 references
Engineering

Abstract

Artificial intelligence (AI) triage systems identify suspected urgent findings on computed tomography (CT) and alert radiologists to support prioritization. Their association with interpretation wait times may vary across clinical settings. We retrospectively evaluated CT examinations acquired between September 2024 and March 2025 and processed by a commercial AI triage platform with nine finding-specific modules. Interpretation wait time was defined as the interval from scan acquisition to report dictation initiation. AI-flagged (AI+) and unflagged (AI−) examinations were compared within emergency department, inpatient, and outpatient classes using medians and 75th percentiles after applying class-specific upper wait-time limits. The analysis included 111,815 examinations, of which 3,846 (3.4%) were AI+. Median wait times for AI+ versus AI− examinations were 11.9 versus 15.3 minutes in the emergency department, 17.5 versus 27.1 minutes for inpatients, and 47.3 versus 1,227.5 minutes for outpatients. These values corresponded to relative differences of 22.2%, 35.4%, and 96.1%, respectively. At the 75th percentile, the corresponding differences were 15.8%, 41.4%, and 56.2%; the outpatient absolute difference was 1,487.8 minutes (24.8 hours). AI-flagged examinations had shorter interpretation wait times in all three settings, with the largest absolute difference among outpatients. These descriptive associations support further controlled evaluation of AI-assisted prioritization but do not establish an AI-attributable reduction in waiting time or improvement in patient outcomes.

View source

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.