Modeling of the geometric variation and analysis of the right atrium and right ventricle motion of the human heart using PCA

Abstract Many heart diseases result in or from abnormal atrium and/or ventricle motion; therefore, analysis of shape changes due to heart motion is crucial to diagnostic classification. This paper aimed at investigating the suitability of a medial-based modeling approach to fulfill the task of representing and analyzing the right atrium and ventricle motion during the cardiac cycle. We analyzed the heart of one subject within a cardiac cycle covering 11 different time steps. Based on shape correspondence of the medial method, a common shape model for the object ensemble was generated for the complete cardiac cycle. The time-dependent variations were analyzed using an extended PCA method showing that almost 90% of variations due to cardiac motion are covered within the first three modes. The promising results highlight the method's value in quantifying shape changes during the cardiac cycle for healthy and for clinical relevant populations.