Clinical Implementation of AI for Pulmonary Embolism Detection in over 30,000 CT Pulmonary Angiography Examinations
Radiology: Artificial Intelligence, 2026
Abstract
Introduction: Acute pulmonary embolism (PE) is a life-threatening condition that requires rapid and accurate diagnosis. Artificial intelligence (AI) tools have been developed to assist radiologists in detecting and triaging PE on CT pulmonary angiography (CTPA), but their real-world performance across large-scale clinical implementations remains an area of active investigation.
Methods: In this large-scale retrospective study, 32,501 real-world CTPA examinations were analyzed to evaluate the concordance between AI-informed radiologists and a U.S. Food and Drug Administration (FDA)-cleared AI tool (Aidoc BriefCase - Pulmonary Embolism Triage) in detecting and triaging acute PE. The study assessed the AI tool's performance in a clinical radiology workflow setting.
Results: Artificial intelligence–informed radiologists demonstrated high concordance with the FDA-cleared AI tool in detecting and triaging acute pulmonary embolism across the 32,501 real-world CT pulmonary angiographic examinations. The findings support the clinical utility of AI-assisted workflows for PE detection in high-volume chest imaging practices.
Conclusion: The integration of AI tools into the radiology workflow for PE detection shows strong agreement with expert radiologist interpretation, supporting the safe and effective clinical deployment of AI-assisted triage systems for pulmonary embolism in real-world settings.
Read the full study: https://doi.org/10.1148/ryai.250017
