Multi-vendor AI across 20 centers speeds trauma radiograph and knee MRI reporting, with data-routing latency the main bottleneck

Radiologist artificial intelligence (AI) adoption hit 91.4%; AI use was tied to 26% lower report turnaround for trauma radiographs and 18% lower for knee magnetic resonance imaging (MRI) (p<0.001). Median latency: 2.06 min, 72% from data routing; 7.2% arrived after finalization.

• Three retrospective cohorts across 20 outpatient centers included 96,874 examinations for latency, 20,909 for report turnaround time, and 58 surveyed radiologists.
• Median total latency was 2.06 minutes (interquartile range 1.74–3.05); data routing accounted for 72%. AI results arrived after report finalization in 7.2% overall (3.0% knee MRI to 13.2% chest CT).
• After radiologist adjustment, AI availability was associated with lower report turnaround time for trauma radiography (-26%) and knee MRI (-18%; both p<0.001), but not brain volumetry MRI (+9.2%; p=0.33); exploratory net promoter score declined for chest CT and aorta CT.

Automated summary

RadPigeon summaries are original and for information only. They are not clinical advice.