Retinal 3D Optical coherence tomography (OCT) is a non-invasive imaging modality in ocular diseases. Due to large volumes of OCT data, it is better to utilize automatic extraction of information from OCT images, such as total retinal thickness and retinal nerve fiber layer thickness (RNFLT). These two thickness values have become useful indices to indicate the progress of diseases like glaucoma, according to the asymmetry between two eyes of an individual. Furthermore, the loss of ganglion cells may not be diagnosable by other tests and even not be evaluated when we only consider the thickness of one eye (due to dramatic different thickness among individuals). This can justify our need to have a comparison between thicknesses of two eyes in symmetricity. Therefore, we have proposed an asymmetry analysis of the retinal nerve layer thickness and total retinal thickness around the macula in the normal Iranian population. In the first step retinal borders are segmented by diffusion map method and thickness profiles were made. Then we found the middle point of the macula by pattern matching scheme. RNFLT and retinal thickness are analyzed in 9 sectors and the mean and standard deviation of each sector in the right and left eye are obtained. The maximums of the average RNFL thickness in right and left eyes are seen in the perifoveal nasal, and the minimums are seen in the fovea. Tolerance limits in RNFL thickness is shown to be between 0.78 to 2.4 μm for 19 volunteers used in this study.
Optical Coherence Tomography (OCT) suffers from speckle noise which causes erroneous interpretation. OCT denoising methods may be studied in "raw image domain" and "sparse representation". Comparison of mentioned denoising strategies in magnitude domain shows that wavelet-thresholding methods had the highest ability, among which wavelets with shift invariant property yielded better results. We chose dictionary learning to improve the performance of available wavelet-thresholding by tailoring adjusted dictionaries instead of using pre-defined bases. Furthermore, in order to take advantage of shift invariant wavelets, we introduce a new scheme in dictionary learning which starts from a dual tree complex wavelet. we investigate the performance of different speckle reduction methods: 2D Conventional Dictionary Learning (2D CDL), Real part of 2D Dictionary Learning with start dictionary of dual-tree Complex Wavelet Transform (2D RCWDL) and Imaginary part of 2D Dictionary Learning with start dictionary of dual-tree Complex Wavelet Transform (2D ICWDL). It can be seen that the performance of the proposed method in 2D R/I CWDL are considerably better than other methods in CNR.