More than 10% of stents will have failure events within a year after intervention, caused mostly be in-stent restenosis or stent thrombosis, and require re-intervention. The non-physiologic biomechanical loads that stents place on coronary arteries have been identified as contributing factors for failure. However, our understanding of the contribution of these factors at the patient-specific level is limited, as computational models lack an understanding of the in vivo deployed-stent geometry. Therefore, the aim of this investigation was to develop a framework to reconstruct this geometry by fusion of intravascular optical coherence tomography (OCT) and micro-computed tomography (µCT) imaging data. Using segmentation techniques on OCT images sparse in vivo stent geometry data is extracted. From a micro-CT scan of a known stent deployment a wireframe model is derived with 15 micrometer voxel resolution. Via diffeomorphic mapping the wireframe is deformed to the sparse OCT-defined geometry, while constraining stent form, resulting in a contiguous 3D reconstructed deployed-stent geometry.
Posted by: Kris Campbell