Cardiorespiratory motion prediction for congenital heart disease
Luc Duong - École de technologie supérieure
Oct. 31, 2025, 2:30 p.m. - Oct. 31, 2025, 3:30 p.m.
ENGMD 279
Hosted by: Kaleem Siddiqi
In this talk, I will introduce our work on cardiorespiratory motion prediction involving 2D, 3D, and 4D analysis of vascular structures in congenital heart disease. Motion prediction is a major issue in interventional cardiology, particularly during the treatment of pediatric congenital heart disease, where the patient’s smaller anatomy and irregular motion patterns present additional challenges. X-ray angiography is used for real-time visualization, but the amount of ionizing radiation must be minimized. These factors directly impact the development of accurate motion evaluation methods. Virtual trials using numerical simulations and virtual phantoms could enable a new generation of techniques, including fully personalized cardiorespiratory motion prediction and 3D models for navigation guidance.
Prof. Duong is Professor at the École de technologie supérieure since 2009. His research focuses on image-guided interventions, medical image analysis, computer vision, and machine learning. He collaborates closely with the cardiology department at CHU Sainte-Justine on projects related to interventions for congenital heart disease. Prof. Duong completed his Ph.D. (2007) at Polytechnique Montréal, on machine learning approaches to guide operative strategies for correcting scoliosis deformities. From 2007 to 2008, he was a postdoctoral fellow at Siemens Corporate Research in Princeton, NJ, USA, working on navigation guidance for coronary arteries with Chronic Total Occlusion (CTO) in collaboration with the Thoraxcentrum in Rotterdam, The Netherlands.
