The emergence of linear array devices employing dry point contact transducers emitting horizontal shear waves significantly increased the efficiency of data acquisition and enabled using imaging techniques for nondestructive evaluation of concrete members. Reverse time migration (RTM) is a mechanics-based imaging technique that has gained the attention of researchers in the context of nondestructive testing (NDT) in recent years. RTM offers a better focusing over synthetic aperture focusing technique (SAFT), a well-established real-time imaging method for NDT of concrete members, and enables locating reflectors with steep slopes and the bottom boundaries of embedded objects. Despite all advantages, RTM suffers from some limitations. It is computationally costly and demands a massive memory. In addition, RTM algorithm generates images with high-amplitude artifacts and assigns amplitudes to the points of a reconstructed image that are not a true representative of the reflectivity of the scanned medium at those points. This dissertation develops an analytical approach to resolve the computational cost and memory demand bottlenecks of the RTM when dry point contact transducers emitting horizontal shear waves are used for data acquisition. Horizontal shear waves preserve more energy than longitudinal waves after emission allowing inspection of concrete members in deeper depths. However, the lower wavelength of shear waves increases the potential of scattering by aggregates and air voids that affects the quality of the reconstructed images. This dissertation develops a 3D numerical tool to study the scattering attenuation of shear waves in concrete. An efficient algorithm is introduced to generate non-overlapping aggregates and air voids to study the effect of size, shape, and material properties of aggregates as well as the porosity of concrete on the scattering attenuation of shear waves. Moreover, it develops novel techniques to mitigate the high-amplitude artifacts of RTM images and to adjust the amplitudes assigned to the points of an image reconstructed by RTM for homogeneous and concrete members.