Three-dimensional modeling and assessment of cardiac adipose tissue distribution.

Objective: The layer of fat that accumulates around the heart, called cardiac adipose tissue (CAT), can influence the development of coronary disease and is indicative of cardiovascular risk. While volumetric assessment of magnetic resonance imaging (MRI) can quantify CAT, volume alone gives no information about its distribution across the myocardial surface, which may be an important factor in risk assessment. In this study, a three-dimensional (3D) modeling technique is developed and used to quantify the distribution of the CAT across the surface of the heart.Methods: Dixon MRI scans, which produce a registered 3D set of fat-only and water-only images, were acquired in 10 subjects for a study on exercise intervention. A previously developed segmentation algorithm was used to identify the heart and CAT. Extracted contours were used to build 3D models. Procrustes analysis was used to register the heart models and an iterative closest point algorithm was used to register and align the CAT models for calculation of CAT thickness. Rays were cast in directions specified by a spherical parameterization of elevation and azimuthal angles, and intersections of the ray with the CAT surface were used to calculate the thickness at each location. To evaluate the effects of the spherical parameterization on the thickness estimates, a set of synthetic models were created with increasing major-to-minor axis ratios.Results: Based on the validation in the synthetic models, the average error in CAT thickness ranged from 1.25% to 17.3% for increasing major-to-minor axis ratio.Conclusions: A process was developed, based on Dixon MRI data, to provide 3D models of the myocardial surface and the cardiac fat. The models can be used in future segmentation algorithm development and for studies on changes in cardiac fat as a result of various interventions.

Empirical Mode Decomposition on Geometry Signal Based Spherical Parameterization

A novel method based onEmpirical Mode Decomposition(EMD) is approached to process the geometry signal. The main idea is to decompose the signal into some different detail components called Intrinsic Mode Function (IMF). The key steps are as follows: First, the signal is spherical parameterization; Second it is transformed into the plane signal and sampled regularly; Third, the translated signal is processed as an image using Bid-Empirical Mode Decomposition, getting several image IMFs; Finally, invert mapping these IMFs to geometry signal and getting the geometry signal’s IMFs.We demonstrate the power of the algorithms through a number of application examples including de-noising and enhancement.

Evaluation of Spherical Parameterization Methods for Three-Dimensional Reconstruction of Medical Images

The spherical parameterization is important for the correspondence problem that is a major part of statistical shape modelling for the reconstruction of patient-specific 3D models from medical images. In this paper, we present comparative studies of five common spherical mapping methods applied to the femur and tibia models: the Issenburg et al. method, the Alexa method, the Saba et al. method, the Praun et al. method and the Shen et al. method. These methods are evaluated using three sets of measures: distortion property, geometric error and distance to standard landmarks. Results show that the Praun et al. method performs better than other methods while the Shen et al. method can be regarded as the most reliable one for providing an acceptable correspondence result. We suggest that the area preserving property can be used as a sufficient condition while the angle preserving property is not important when choosing a spherical mapping method for correspondence application.