According to ARRS’ American Journal of Roentgenology (AJR), an AI tool for detection of incidental pulmonary embolus (iPE) on conventional contrast-enhanced chest CT examinations had high NPV and moderate PPV, even finding some iPEs missed by radiologists

Left: Axial image shows central nonocclusive filling defect in right lower lobe lateral subsegmental pulmonary artery (arrow). Examination classified positive for iPE by AI but negative for iPE by clinical report. Adjudication process used for reference standard determination classified examination as positive for iPE. Right: Activation map shows location corresponding with iPE identified by AI (arrow). 


July 21, 2022 — According to ARRS’ American Journal of Roentgenology (AJR), an AI tool for detection of incidental pulmonary embolus (iPE) on conventional contrast-enhanced chest CT examinations had high NPV and moderate PPV, even finding some iPEs missed by radiologists. 

“Potential applications of the AI tool include serving as a second reader to help detect additional iPEs or as a worklist triage tool to allow earlier iPE detection and intervention,” wrote lead investigator Kiran Batra from the University of Texas Southwestern Medical Center in Dallas. “Various explanations of misclassifications by the AI tool (both false positives and false negatives) were identified, to provide targets for model improvement.” 

Batra and colleagues’ retrospective study included 2,555 patients (1,340 women, 1,215 men; mean age, 53.6 years) who underwent 3,003 conventional contrast-enhanced chest CT examinations between September 2019 and February 2020 at Parkland Health in Dallas, TX. Using an FDA-approved, commercially available AI tool (Aidoc, New York, NY) to detect acute iPE on the images, a vendor-supplied natural language processing algorithm (RepScheme, Tel Aviv, Israel) was then applied to the clinical reports to identify examinations interpreted as positive for iPE. 

Ultimately, the commercial AI tool had NPV of 99.8% and PPV of 86.7% for detection of iPE on conventional contrast-enhanced chest CT examinations (i.e., not using CT pulmonary angiography protocols). Of 40 iPEs present in the team’s study sample, 7 were detected only by the clinical reports, and 4 were detected only by AI. 

Noting that both the AI tool and clinical reports detected iPEs missed by the other method, “the diagnostic performance of the AI tool did not show significant variation across study subgroups,” the authors of this AJR article added

For more information: arrs.org 

Related Content

News | Magnetic Resonance Imaging (MRI)

August 18, 2022 — According to ARRS’ American Journal of Roentgenology (AJR), machine learning models applied to ...

Time August 18, 2022
arrow
Sponsored Content | Videos | Artificial Intelligence

Digital solutions like AI and the cloud are driving innovation and transforming the way radiologists and imaging centers ...

Time August 18, 2022
arrow
News | X-Ray

August 17, 2022 — The Institute of Human Virology in Nigeria (IHVN) has announced their partnership with Fujifilm, the ...

Time August 17, 2022
arrow
News | Computed Tomography (CT)

August 15, 2022 — According to ARRS’ American Journal of Roentgenology (AJR), the combination of deep-learning ...

Time August 15, 2022
arrow
News | Ultrasound Imaging

August 10, 2022 — Butterfly Network, Inc., a digital health company transforming care with handheld, whole-body ...

Time August 10, 2022
arrow
News | Lung Imaging

August 8, 2022 — Representatives from the Diagnostics Working Group of the IASLC Early Detection and Screening Committee ...

Time August 08, 2022
arrow
News | Computed Tomography (CT)

August 4, 2022 — Xoran Technologies announces that last month it received FDA 510(k) clearance for TRON — a truly mobile ...

Time August 04, 2022
arrow
News | Contrast Media

August 3, 2022 — According to an open-access article in ARRS’ American Journal of Roentgenology (AJR), electronic health ...

Time August 03, 2022
arrow
News | Contrast Media

August 2, 2022 – A new update has been announced as the radiology world continues to address supply disruptions of ...

Time August 02, 2022
arrow
Feature | Radiology Business | By Melinda Taschetta-Millane

Here is what you and your colleagues found to be most interesting in the field of medical imaging during the month of ...

Time August 01, 2022
arrow
Subscribe Now