Robust physically-constrained modeling of the mitral valve and subvalvular apparatus

Area of Science:

• Cardiovascular Imaging
• Biomedical Engineering
• Computational Biology

Background:

• The mitral valve (MV) apparatus is complex and frequently implicated in cardiac diseases.
• Accurate, patient-specific modeling of the dynamic MV is challenging due to its intricate shape and motion.

Purpose of the Study:

• To develop a novel, personalized modeling approach for the dynamic mitral valve and subvalvular apparatus.
• To ensure temporal consistency and realistic deformations in cardiac sequences.
• To provide discriminant biomarkers for detecting and quantifying cardiac remodeling.

Main Methods:

• A two-step alternate algorithm combining discriminative learning and leaflet biomechanics.
• Detection of anatomical MV components under biomechanical constraints.
• Evaluation on transesophageal echocardiographic sequences and clinical subject data.

Main Results:

• Achieved an average Hausdorff error of 5.1 mm at 9 sec, an 11.5% improvement over data-driven methods.
• Demonstrated the fully-automatic approach's ability to provide discriminant biomarkers.
• Successfully detected and quantified annulus and leaflet remodeling in functional mitral regurgitation.

Conclusions:

• The proposed method offers a robust and accurate approach for personalized mitral valve modeling.
• This technique enhances the understanding and quantification of cardiac remodeling in mitral regurgitation.
• The automated system provides valuable biomarkers for clinical assessment.