Application of Complex Wavelet Transform in OCT Image Compressive Sensing Reconstruction

This paper presents a method for reconstructing Optical Coherence Tomography (OCT) images from incomplete data, a crucial advancement for mitigating acquisition time and motion artifacts. For this reason, the cross-sectional OCT data is sampled using different sampling strategies for the acquisition data and different algorithms are employed to reconstruct the high SNR and high-resolution OCT data in two and three dimensions. The reconstruction problem is solved by using a two-phase approach, including sparse coding and dictionary updating. At first, the sparsity of data is investigated by Discrete Cosine Transform (DCT), wavelet and complex wavelet transforms as initial dictionary. Then, the missing data is reconstructed by a compressive sensing (CS) reconstruction approach. Simulation results demonstrate that the dictionary learned from the complex wavelet transform yields the sparsest image representation, leading to improved CS reconstruction. The proposed CS reconstruction algorithm offers the potential to significantly accelerate OCT data acquisition while reducing the impact of motion-induced distortions, representing a valuable contribution to the field of OCT imaging.