Estimation of dynamically evolving ellipsoids with applications to medical imaging

Area of Science:

• Medical Imaging
• Computational Geometry
• Biomedical Engineering

Background:

• Left-ventricular ejection fraction (LVEF) is a critical indicator of cardiac function.
• Current LVEF estimation methods often rely on specific imaging modalities and can be sensitive to noise.
• Dynamic estimation of LVEF provides a more comprehensive assessment of cardiac performance.

Purpose of the Study:

• To develop and evaluate a model-based approach for dynamic LVEF estimation.
• To utilize geometric reconstruction and recursive estimation techniques for improved accuracy.
• To address limitations of traditional LVEF estimation methods using myocardial perfusion imaging.

Main Methods:

• Modeled the left ventricle as a dynamically evolving 3D ellipsoid.
• Represented the left-ventricular outline as a 2D elliptical projection from gated planar myocardial perfusion images.
• Employed a Rauch-Tung-Striebel smoothing filter for robust ellipsoid reconstruction and dynamic LVEF estimation.
• Investigated both 3D reconstruction and combined 2D ellipse estimation approaches.

Main Results:

• The proposed method provides a dynamic LVEF estimate that is more robust to noise by utilizing the entire dataset.
• Numerical studies demonstrated the sensitivity of the approach to dynamics and projection geometry.
• The method was validated using both simulated and real myocardial perfusion imaging data.

Conclusions:

• The model-based approach offers a robust and dynamic method for LVEF estimation from myocardial perfusion images.
• This technique enhances LVEF assessment by integrating geometric reconstruction and recursive estimation.
• The findings provide a basis for specifying system parameters in dynamic cardiac function analysis.