title:: Complete valvular heart apparatus model from 4D cardiac CT
publisher:: Elsevier
people:: Sasa Grbic, Razvan Ionasec, Dime Vitanovski, Ingmar Voigt, Yang Wang, Bogdan Georgescu, Nassir Navab, Dorin Comaniciu
organization:: Siemens Corporate Research, Technical University of Munich
domain:: Medical Imaging, Machine Learning, Computational Modeling, Cardiology
link:: https://doi.org/10.1016/j.media.2012.02.003
citations:: 65 (OpenAlex) · 29 (Semantic Scholar) — as of 14 October 2026
First page of the paper — author copy hosted by the FAU Pattern Recognition Lab. Corresponding-author contact details redacted.
Summary
Valvular heart disease rarely involves one valve in isolation, and it rarely gets treated that way. This paper extends patient-specific modelling from the left-heart valves to the complete valvular apparatus - aortic, mitral, pulmonary and tricuspid - as one modular model estimated from 4D cardiac CT.
The technical contribution is a constrained Multi-linear Shape Model (cMSM), conditioned by anatomical measurements, that represents the complex spatio-temporal variation of all four heart valves together. The cMSM sits inside a learning-based framework that estimates patient-specific valve parameters efficiently from cine images. Combined with existing patient-specific chamber models, this is what makes personalised computational modelling and realistic simulation of the entire cardiac system possible.
Data Points
- All four heart valves - aortic, mitral, pulmonary, tricuspid - in one modular patient-specific model
- New constrained Multi-linear Shape Model (cMSM) conditioned by anatomical measurements
- Represents the spatio-temporal variation of the complete valvular apparatus, not one valve at a time
- Evaluated on 64 4D cardiac CT studies
- Enables automatic quantitative evaluation of the full valvular apparatus from non-invasive imaging
- Together with chamber models, the geometric basis for whole-heart simulation
Reference
Medical Image Analysis 16(5):1003-1014, 2012. doi:10.1016/j.media.2012.02.003 · PubMed 22481023
65 citations (OpenAlex, October 2026)
Part of the complete record in Scientific Publications - Peer-Reviewed Research Record.
