.jpg)
SIR 2025
Venous Interventions
Educational Exhibit
Mary E. McLaughlin, MD
PGY-5 IR/DR Resident
Weill Cornell Medical Center, United States
Eli G. Gussen, III, MD, MS
Research Coordinator
Weill Cornell Medical Center, United States
Christopher Harnain, MD, MBA, RPVI
Vascular & Interventional Radiologist
Division IR Weill Cornell Medical Center, United States
1. Introduce the clinical utility of artificial intelligence (AI) through an ongoing example.
2. Emphasize how AI can mitigate the challenges of screening large patient populations.
3. Detail our experience in how AI can contribute as an additional member of the pulmonary embolism response team (PERT).
Background: A challenge of providing appropriate clinical treatment of patients with pulmonary embolism (PE) is the identification of patients who might benefit from catheter-directed therapy. This challenge is particularly relevant today, as expanding treatment options, which are often not well known to referring providers, become more readily available with advancing technology. At our institution, we have addressed this challenge through the use of an AI screening application.
Clinical Findings/Procedure Details: Starting in November 2023, we have utilized Viz.AI, an artificial intelligence application (VizRecruit version 1.93.10 application provided by Viz.AI San Francisco, CA), to screen all computed tomography chest exams at Weill Cornell Medicine in order to identify patients with main or lobar PE and measured RV/LV ratio > 1.0. Subsequently, the AI-identified patients were evaluated by the interventional radiology team for consideration of catheter-directed therapy, in addition to anticoagulation. The use of this AI application in our clinical practice has enabled earlier detection of PE and efficient screening of a large population of patients, without relying on formal consultation through PERT notifications alone. In this way, the AI application has not only allowed for increased enrollment in ongoing clinical trials, but it has also led to improvement in overall patient care.
Conclusion and/or Teaching Points: In conjunction with traditional PERT notifications through Epic, the use of this AI application at our institution has allowed for earlier detection of PE, evaluation of more patients who might benefit from catheter-directed therapy, and enhanced screening for enrollment in ongoing clinical trials.