Investigators across Europe leverage PRAEVAorta2 AI-driven segmentation on pre- and post-EVAR CT angiograms, combining imaging and clinical variables in deep learning models to forecast postoperative outcomes and optimize surveillance strategies for aortic aneurysm patients.
Key points
Automated segmentation and morphometric measurement of aneurysms using CE-marked PRAEVAorta2 on CT angiography
Integration of clinical, procedural, and imaging features into deep convolutional neural networks for postoperative risk stratification
Multicenter retrospective cohort of 500 EVAR patients with 70/30 training-testing split to develop and validate predictive models
Why it matters:
This protocol establishes AI-enabled precision surveillance and risk stratification post-EVAR, potentially reducing complications and personalizing vascular care.
Q&A
What is EVAR?
What are endoleaks and why do they matter?
How does PRAEVAorta2 work?
What is a retrospective cohort study?
Why split data into 70% training and 30% testing sets?
Read full article
Academy
Artificial Intelligence in Aortic Aneurysm Imaging
Overview: Abdominal aortic aneurysms (AAA) are bulges in the major artery supplying blood to the abdomen and legs. If untreated, they can rupture and cause life-threatening bleeding. Monitoring AAA size and shape over time is essential to decide when to intervene. Computed tomography angiography (CTA) is the standard imaging method, providing detailed cross-sectional images of blood vessels.
Traditionally, radiologists manually trace the aneurysm on each CT slice to measure its diameter and volume. This process is time-consuming, variable between operators, and limits the size of studies. Recent advances in artificial intelligence (AI) and deep learning offer automated solutions to segment the aneurysm and extract morphometric data rapidly and consistently.
What Is Automated Segmentation?
- Automated segmentation uses computer algorithms to identify and outline anatomical structures in medical images without manual input.
- Deep neural networks, inspired by the human brain, learn features from thousands of labeled images to distinguish vessels from surrounding tissue.
- Once trained, the algorithm can process new CT scans to produce precise contours of the aneurysm wall, lumen, thrombus, and calcifications.
Key Benefits for Patients
- Speed: Automated workflows analyze scans in minutes instead of hours, enabling faster clinical decisions.
- Consistency: AI models apply the same criteria across all cases, reducing variability.
- Detail: Advanced algorithms can measure complex features such as vessel angulation and tortuosity that are hard to quantify manually.
How AI Models Are Trained
NIH and academic centers supply large libraries of anonymized CT scans. Expert radiologists manually segment a subset of these images to create a ‘ground truth.’ The AI model adjusts its internal parameters to minimize the difference between its output and the ground truth. This training phase uses hundreds to thousands of examples. Model performance is then tested on new cases not seen during training.
Clinical Implementation
CE-marked software like PRAEVAorta2 has cleared regulatory review for clinical use in Europe. Hospitals integrate AI tools into their PACS (picture archiving and communication system) workflows. After EVAR, clinicians compare baseline and follow-up scans processed by AI to detect changes in aneurysm size or complications such as endoleaks. This data guides personalized follow-up intervals, reducing unnecessary scans and focusing care on high-risk patients.
Relevance to Longevity and Healthspan
By improving surveillance after aneurysm repair, AI-driven imaging helps maintain vascular health, prevent catastrophic events, and support longer, healthier lives. Automated analysis frees physician time for patient care and allows large-scale studies to uncover factors affecting long-term outcomes.