Ph.D Thesis

Ph.D StudentLysiansky Michael
SubjectTwo Dimensional Ultrasound Speckle Tracking and Inverse
Acoustic Scattering Problem
DepartmentDepartment of Biomedical Engineering
DR. Michael Zibulevsky
Full Thesis textFull thesis text - English Version


1.                 Two Dimensional Ultrasound Speckle Tracking

Quantitative accurate reproducible analysis of left ventricular (LV) regional wall motion is of major importance in patients with diseased coronary circulation. The early clinical manifestation of these pathologies is myocardial mechanical dysfunction at various levels.

In this study, the new method for the strain analysis of two-dimensional cross-sections of LV has been developed, tested and analyzed. The method is based on generalization of 2D Strain algorithm, which so far has been successfully applied and clinically validated. The new method of 2D strain analysis provides more detailed information regarding the motion of the LV wall. It allows evaluation of the transmural strain variation and introduces new parameters, describing deformation field. These parameters have potential clinical value, especially in diagnostic of non-transmural pathologies.

The developed tracking algorithm is robust with respect to speckle noise. The influence of the speckle noise was reduced using both smoothness of the deformation field and the periodicity of the deformations in time. The performance of the new algorithm was validated on simulated data. The algorithm was applied to several of ultrasonic clips of normal subjects and different patients. The behavior of the different parameters, characterizing the deformation field, was analyzed.  It was demonstrated that the measured strain values are consistent with the results obtained using 2D Strain algorithm and MRI.

Altogether, our results show that the generalized 2D Strain algorithm is a novel method enabling cardiac muscle deformations studies, for life sciences research and clinical applications.

2.                 Inverse acoustic scattering problems

Medical ultrasound imaging is widely used for obtaining information about the anatomical structure and the physiological function of organs within the human body. The propagation of ultrasound waves in soft tissue is characterized by three different functions of the spatial coordinates: density, compressibility and absorption. The ultrasonic image is a graphical representation of a combination of these 3 functions, which is called the tissue reflectivity function (TRF). The estimation of the TRF from the ultrasound waves, echoed back from the tissue, and its relationship to tissue properties, have been addressed by many studies but still remain as open questions. Here, a method of restoration of the TRF from a scattered data is presented, for the case of frequency independent absorption and the presence of only weak diffuse scatters. The method was validated on the numerically simulated phantoms, and was found to accurately restore the ‘tissue’ impedance changes.