Three dimensional ultrasound image reconstruction using nonlinear wavelet thresholding

Two-dimensional (2D) noisy ultrasound image results in low soft tissue contrast, thus limiting the effectiveness of three dimensional (3D) imaging. This paper presents nonlinear wavelet thresholding methods for noise reduction and thus improves the robustness of 3D reconstruction. A sequence of 60 slices with 256×256×8 ultrasound images of a Ping-Pang ball in a water tank have been utilized for 3D ultrasound reconstruction. Experimental results show that the distinct features of 3D reconstructed object can be obtained from both 60 and 40 slice images after applying nonlinear wavelet threshold for noise reconstruction