Patient-specific numerical simulations of endovascular procedures in complex aortic pathologies
Objective
Endovascular repair has become the first-line treatment for many complex aortic pathologies, including abdominal, thoracic and thoraco-abdominal aneurysms as well as aortic dissections.
However, its clinical success depends heavily on selecting the right stent-graft for each individual patient and on the operator's experience. Today's planning tools β mostly based on CT-scan reconstruction β cannot predict how a device will actually behave inside a specific anatomy.
The objective of this review is to introduce the basic principles of numerical simulation and to assess the current literature on its use for the endovascular management of complex aortic disease, showing how patient-specific computer modeling could transform the way these interventions are planned and performed.
Methods
We conducted a narrative review of patient-specific numerical simulation applied to endovascular procedures.
It is structured around the three stages of any simulation workflow: preprocessing (segmenting CT or MRI images into a patient-specific 3D model and assigning material properties and boundary conditions), solving, and post-processing the results into clinically usable information.
We surveyed the published evidence organized by aortic pathology β EVAR for abdominal aneurysms, TEVAR for thoracic aneurysms, fenestrated and branched devices (FEVAR) for juxta-renal and thoraco-abdominal aneurysms, and aortic dissections.
We also examined the main modeling approaches: finite element structural analysis, computational fluid dynamics, fluid-structure interaction, and emerging machine-learning and digital-twin methods.
Results
Finite element analysis is used in the large majority of studies to predict stent-graft deployment, using virtual catheter, virtual shell or direct placement techniques. Computational fluid dynamics is applied to study blood flow and wall shear stress, particularly in dissections.
Validated clinical applications are emerging: a finite element study of 51 FEVAR patients predicted target-artery positions with 95% of measurements within 3 mm and 15Β° of the reference.
A prospective six-center study of 50 patients confirmed satisfactory accuracy and reduced device delivery time β with results now used by manufacturers in device design.
Current limitations include approximations of tissue material properties, long computation times, and a lack of large-scale prospective validation.
Conclusion
Numerical simulation already addresses many of the key challenges in the endovascular management of complex aortic disease, and for several of them the models are close to being ready to help clinicians choose the most tailored solution for each patient.
Although the technology is still in its infancy for the most complex devices and needs broader validation before routine use, this review shows that patient-specific simulation is about to revolutionise how we plan and carry out endovascular interventions.