14493
clinical study

AI-Driven Triage of Intracranial Hemorrhage on ED Head CT Improves Radiologist Turnaround Time

Materials & Methods
All noncontrast Emergency Department (ED) head CTs receiving AI-positive intracranial hemorrhage (ICH) notifications across a three-hospital integrated system over 12 months were reviewed. Overnight cases (10 p.m.–7:30 a.m.) were excluded due to lack of AI access. Studies were categorized as true positive, false positive or follow-up of known ICH. Turnaround time (TAT) was measured from scan availability to first radiologist access. TAT for new ICH cases was compared with FP and negative studies.

Results
Of 14,707 ED head CT examinations, 632/14,707 cases (4.3%) generated AI-positive notifications, of which 383/632 (61%) were true positives and 243/632 (39%) were false positives; 54/632 true positive cases (8%) represented follow-up studies. The mean TAT for new ICH cases was 12.9 minutes, compared with 16.2 minutes for false positive cases and 15.7 minutes for AI negative cases. Differences in TAT between new true positive and false positive cases were statistically significant (p = 0.02) and became more pronounced after excluding follow-up examinations (p = 0.001). Radiologists triaged examinations using PACS-integrated AI widgets that displayed key images and status indicators, enabling rapid case prioritization.

Conclusions
AI-based ICH notifications significantly reduced TAT for newly identified hemorrhages, demonstrating that radiologists leverage AI alerts to prioritize high-risk cases. Faster access to new ICH studies may improve time-to-diagnosis, enhance emergency stroke workflows and positively impact patient outcomes.

Disclaimer: This links to a third-party website, which may require credentials to review.

Explore the Latest AI Insights, Trends and Research