14490
clinical study

Boosting Pulmonary Embolism Detection With FDA-Cleared AI: Uncovering Hidden Cases in a Retrospective Study

Incidental Findings, Incidental Pulmonary Embolism, Pulmonary Embolism

Materials & Methods
This retrospective study evaluated an FDA-cleared artificial intelligence (AI) imaging platform (Aidoc) for detecting symptomatic and incidental pulmonary embolism (PE) on CT imaging. A total of 2,349 CT pulmonary angiograms (CTPAs) and 4,609 CT chest, abdomen, and pelvis (CTCAP) or abdominopelvic CT (CTAP) examinations performed for non-PE indications between September 24, 2023, and November 30, 2023, were analyzed. Natural language processing (NLP) was used to classify radiology reports as PE-positive or PE-negative, while the AI algorithm analyzed imaging data. Cases flagged as PE-positive by AI but negative by NLP were considered discordant and independently reviewed by a board-certified radiologist to determine true positives (TPs), false positives (FPs), or questionably positive (QPs). Enhanced detection rate was calculated as the proportion of AI-detected TPs missed by radiologists relative to radiologist-detected positives.

Results
Radiologists initially identified 142 PEs and 32 incidental pulmonary embolisms (iPEs), all of which were confirmed by AI at concordant anatomical locations. The AI flagged 32 (1.4%) CTPAs and 15 (0.3%) CTCAP/CTAP studies as discordant. Expert review confirmed 17/32 (53%) AI-flagged PE cases as true positives, 9 (28%) as false positives, and 6 (19%) as questionably positive. Among incidental PE cases, 12/15 (80%) were confirmed as true positives, 1 (7%) as false positive, and 2 (13%) as questionably positive. Overall, AI increased detection by 12% (17/142) for pulmonary embolisms and by 38% (12/32) for incidental pulmonary embolisms compared with radiologist interpretation alone. Expanded chart review demonstrated that several patients with additional AI-detected findings were already receiving anticoagulation therapy, suggesting minimal downstream clinical harm.

Conclusions
The FDA-cleared AI tool enhanced detection of both pulmonary embolism and incidental pulmonary embolism by identifying cases missed during routine radiologist interpretation. Most AI-detected discordant findings were confirmed as true positives upon expert review. The limited downstream clinical impact further supports its safe integration into clinical workflows. 

VIEW CONFERENCE

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

Explore the Latest AI Insights, Trends and Research

Follow a manual added link

Webinars

Aidoc experts, customers and industry leaders share the latest in AI benefits and adoption.

Follow a manual added link

eBooks

Explore how clinical AI can transform your health system with insights rooted in real-world experiences.

Follow a manual added link

AI Strategy

Learn how to go beyond the algorithm to develop a scalable AI strategy and implementation plan.

Follow a manual added link

Solutions

Explore how Aidoc can help increase hospital efficiency, improve outcomes and demonstrate ROI.

From Our Blog

View all

© Copyright 2026 Aidoc