Optical Coherence Elastography Based On Wavelet Theory

In this thesis, we present an optical technique to image elastic properties of the human thyroid with suspected cysts utilizing optical coherence elastography (OCE). The influences of kernel size and the dilation parameter on calculations were investigated via an experimental agarose phantom to verify the feasibility and stability of OCE in advance.Firstly, internal displacements were calculated by2D normalized cross-correlation, and the corresponding displacement vectors were described graphically. Then the continuous wavelet transform was used as a more effective and stable method to improve the accuracy of estimation. Four types of images including axial/lateral displacement map, axial strain map, modulus distribution and Poisson’s ratio images were obtained to characterize elasticity variations of the phantom. Results suggest that the presented system enables to detect the shape and position of the inclusion, and distinguish the inclusion from background. In addition, elastic properties images of thyroid with suspected cysts were depicted to distinguish benign lesions from malignancy qualitatively. Thus, elastic properties imaging based on OCE shows great promise for the detailed characterization of lesions and preliminary diagnosis of human thyroid diseases as a supplementary method.