Characterization of the heart muscle aniosotropy using ultrasound Nakagami imaging

Abstract

The myocardial architecture of the left ventricle is important for diagnosis, as it is associated with the mechanical and electrical functions. Nakagami imaging has been used in modeling the statistics of backscattered ultrasound signals in soft tissues, and the m value is found capable of differentiating various scattering conditions. Therefore, this study aims to apply Nakagami imaging to explore its feasibility in evaluation of the structural anisotropy in the myocardium. Three excised porcine left ventricles were embedded in gelatin phantoms and imaged in standard long- and short-axis views using the coherent plane wave compounding technique and a linear array probe operating at 5.2 MHz. The Nakagami parametric map was generated using a window-based method using an optimized window size of 12λ × 12λ. Four parameters, including mean, variance, skewness and kurtosis, were used to characterize the distribution of m values. Our results showed that the distributions of m values in the middle layers were statistically different from the subepicardial and subendocardial layers (p < 0.05). The mean m values were significantly different both between anterior and posterior heart walls and between long- and short-axis views (p < 0.05). Our preliminary results showed a potential of Nakagami imaging in myocardial anisotropy characterization. © 2014 IEEE.