Motion compensation during percutaneous interventions This research axis focuses on the characterization of motion during percutaneous interventions using image-based techniques. Our goal is to use novel computational methods to enhance typical X-ray angiographic sequences with temporal reconstruction and provide proper motion modelling using artificial intelligence and machine learning.
Computational Fluid Dynamics (CFD) simulations for stent design The choice of stent and it's placement in vascular structures is a complex task. This research axis aimed at understanding and simulating stent placement, including automatic stenosis detection, considers foreshortening, which is the percentage of reduction of stent length due to expansion, and was validated using annotations by cardiologists before and after stenting and against in vivo pressure measurements. Personalized approaches to design stents would be powerful tools to interventional cardiologists to treat stenosis in vascular structures. For more info: https://substance.etsmtl.ca/en/concevoir-stents-demain-cardiopathies-congenitales
Intravascular imaging of coronary arteries This research axis is aimed towards a better characterization of coronary tissues from OCT imaging. Our team proposed using deep learning feature extraction to analyse coronary arteries tissue (intima, media, adventicia). Our approach identify automatically tissues and register frame-by-frame for a 3D representation. For more info : https://technologytransfer.ca/en/portfolio-posts/fully-automatic-diagnostic-model-of-coronary-artery-lesions-using-optical-coherence-tomography-oct-imaging/