Ontological Labels for Automated Location of Left Ventricular Remodeling

In a previous study, Ardekani et. al. used the Large Deformation Diffeomorphic Metric Mapping (LDDMM) image registration algorithm to analyze the left ventricle of 25 human subjects at both end systole (ES) and end diastole (ED) phases of the cardiac cycle. Of these 25 subjects, 12 had nonischemic cardiomyopathy (NICM) and 13 had ischemic cardiomyopathy (ICM). An average template image at each of ES and ED was generated and then mapped using LDDMM to each patient image. A statistical analysis of the voxel-based Jacobian map, which encodes local tissue volume difference, found one region of statistically significant volume expansion in the NICM group compared to the ICM group. Visual inspection by an expert clinician located this region in the mid anterior subregion of the left ventricle. Using vSPARQL-an extension of SPARQL allowing for recursive and sub-queries the subclass hierarchy surrounding the term Region_of_mγocardium was extracted from the Foundational Model of Anatomy (FMA).

I. Introduction

In a previous study, Ardekani et. al. [1] used the Large Deformation Diffeomorphic Metric Mapping (LDDMM) [2] image registration algorithm to analyze the left ventricle of 25 human subjects at both end systole (ES) and end diastole (ED) phases of the cardiac cycle. Of these 25 subjects, 12 had nonischemic cardiomyopathy (NICM) and 13 had ischemic cardiomyopathy (ICM). An average template image at each of ES and ED was generated and then mapped using LDDMM to each patient image. A statistical analysis of the voxel-based Jacobian map, which encodes local tissue volume difference, found one region of statistically significant volume expansion in the NICM group compared to the ICM group. Visual inspection by an expert clinician located this region in the mid anterior subregion of the left ventricle.