PE AI study shows high sensitivity but need for radiologist oversight
A Northwell Health study found 97.8% agreement between radiologists and an AI tool for pulmonary embolism detection on CT pulmonary angiography.
A radiologist-AI workflow showed 99.2% sensitivity for pulmonary embolism detection in a large real-world CT pulmonary angiography study, according to the Harvey L. Neiman Health Policy Institute.
Researchers from Northwell Health analyzed 32,501 CTPA exams performed across an integrated healthcare system over an 18-month period. Overall agreement between the Aidoc AI tool and radiologist interpretations was 97.8%.
Concordance was higher for negative exams than positive exams, at 98.18% versus 93.75%, respectively. The Neiman Institute said the finding points to the algorithm’s strength in helping rule out pulmonary embolism.
Discordant cases were independently reviewed by expert thoracic radiologists. In those disagreements, radiologists were correct in 88.7% of cases, while AI was correct in 11.3%.
Among confirmed pulmonary embolism cases, 15% were identified by radiologists but missed by the AI tool alone. Matthew Barish, MD, said the findings showed that “radiologist oversight remained necessary.”
The AI system flagged suspected positive CTPA cases to assist radiologists with triage inside the Northwell Health workflow. Agreement was highest for acute and central emboli, the cases associated with greater clinical urgency and mortality risk, according to the release.
Pulmonary embolism is responsible for 5% to 10% of in-hospital deaths and more than 300,000 deaths annually in the U.S., the Neiman Institute noted. The study authors said few large-scale evaluations have assessed FDA-cleared PE detection tools using a human-in-the-loop adjudication model in routine clinical practice.
The results were published in Radiology: Artificial Intelligence.
Sources
About the author
RadiologySignal.com writersEditorial Team
Radiology Signal Staff covers developments across medical imaging, radiology AI, imaging informatics, clinical research, and radiology business. The team monitors primary sources, peer-reviewed studies, company announcements, society updates, and healthcare industry news to deliver concise reporting for imaging professionals.
More from this section
SimonMed highlights role in large global X-ray AI study
The study analyzed 258,373 X-rays from 100 medical centers across 26 countries and evaluated the Rayvolve AI Suite across trauma, chest, measurement, and bone age use cases.
Former RadNet CTO joins Neurophet as scientific advisor
The neuroradiologist and former RadNet CTO will advise the neuroimaging AI company as it seeks to strengthen clinical validation and U.S. market expansion.
CMR-CLIP AI model interprets cardiac MRI scans with high accuracy
The model was trained on more than 13,000 cardiac MRI studies and linked moving heart images with clinical radiology reports, according to the research team.
Flywheel-AWS link creates governed imaging layer for trials
The integration combines AWS HealthImaging’s cloud-based DICOM storage with Flywheel’s imaging data orchestration tools for clinical trials, research, and AI development.
Free HelloAI training targets practical healthcare AI adoption
The company is making the HelloAI Professional track available at no cost, adding more than 20 hours of on-demand healthcare AI education.
Siemens links coronary CT planning to cath-lab PCI guidance
The system combines Syngo.CT Coronary Cockpit and Syngo PCI Connect to transfer preprocessed coronary CT angiography data into the cath lab.
AZmed gets FDA clearance for expanded AZtrauma scope
The clearance expands AZtrauma beyond fracture detection to include joint effusions and dislocations on X-rays in adults and children ages 2 years and older.
HOPPR breast imaging model converts 2D mammograms into text
The HOPPR EB 2D Mammo Narrative Model generates structured narrative output from 2D mammography images for breast imaging workflow applications.
DeepHealth secures AI clearances for neuro, prostate tools
The regulatory milestones cover Neuro Suite Brain Health and Brain Age, LumbarMR, and Prostate AI within DeepHealth’s clinical AI portfolio.
Philips SmartIQ aims to cut coronary imaging radiation dose
The coronary imaging technology adds an ultra-low-dose protocol for Philips’ Azurion image-guided therapy platform and is available in Europe and selected markets.
Azra AI, RevealDx partner on lung nodule AI
Azra AI will integrate RevealAI-Lung into its incidental findings and oncology workflow platform.
Accuro XV clearance gives Rivanna path into MSK ultrasound
Rivanna received FDA 510(k) clearance for Accuro XV, a diagnostic ultrasound system for musculoskeletal imaging. The company is also developing AI tools for bone visualization and fracture detection.