dynamic programming

Retina is the innermost tissue in human eye which sends visual information to the brain by means of optic nerve. Designing an intra-retinal layer segmentation method which can detect the retina surfaces properly in the presence of noise and lack of contrast is an important step in ophthalmology. In this study the combination of morphological component analysis (MCA) and dynamic programming (DP) is used automatically for segmentation of optical coherence tomography (OCT) images.
Data set for this study was 55 samples which were taken from normal people by Topcon OCT-1000. This study had two phases. In MCA phase the image dictionary was created by clustering with eigenvalues (k-SVD), and then the image was decomposed to cartoon and texture parts by selecting proper bases. In the second phase segmentation was done by the dynamic programming (DP) method on cartoon part and the retinal pigment epithelium (RPE), Verhoeff's memberane (VM), outer segment layer (OSL), inner collagenous layer (ICL), inner synaptic layer (ISL), and outer limiting membrane (OLM) layers were detected.
Findings: Comparing the obtained results with gold standard (manual segmentation) shows that minimum error belongs to OSL surface and its error in the form of mean ± SD (standard derivation) is 0.030 ± 0.167. For other surfaces the error is calculated in this way from left to right for RPE, VM, ICL, ISL, OLM: 0.66 ± 0.33, -0.59 ± 0.31, -1.00 ± 0.49, -1.72 ± 0.61, -1.05 ± 0.51.
MCA in combination with DP can work as an automatic method for six lower intra retina layer's segmentation with acceptable accuracy. One of the main advantages of this method is omitting preprocessing phase for segmentation.

Enhanced depth imaging optical coherence tomography images (EDI-OCT) is used for detailed imaging of the choroid layer that contains the highest amount of blood flow in the eye and is affected in several diseases such as choroidal polyps, age-related degeneration and central serous chorioretinopathy. Choroidal segmentation is really important, but the manual segmentation is time consuming and encounters difficulties when large numbers of data is available. Since a large amount of information is available in the images, non-automated and visual analysis of data is almost impossible for the ophthalmologist. The main goal of automatic segmentation was to help the ophthalmologists in the diagnosis and monitoring diseases related to the eye.

The data used in this project was obtained from the Heidelberg OCT-HRA2-KT instrument. Fifty 2 dimensional data were used to evaluate the algorithm. In this study, the retinal pigment epithelium (RPE) and choroid was segmented using a boundary detection algorithm named dynamic programming.

The proposed algorithm was compared with the manual segmentation and the results showed an unsigned error of 1.71 ± 0.93 pixels for retinal pigmented epithelium (RPE) extraction and 10.48 ± 4.11 pixels for choroid detection. It showed significant improvements over other approaches like k-means method.

A few automated methods are applied in the choroid segmentation and most of the studies were mainly focused on the manual separation. In this study, a fast and automated method was provided for the segmentation of choroid area.