Formation and Fusion of Projection Images from 11 Layers of Retina Using Statistical Indicators to Obtain an Image with Appropriate Contrast from the Retinal Depth

Optical coherence tomography (OCT) is a progressively important modality for the noninvasive management of retinal diseases، including age-related macular degeneration (AMD)، glaucoma، and diabetic macular edema. Spectral domain OCT (SD-OCT) generates 3-dimensional (3-D) volumes، which have proven to be useful in clinical practice. In this regards، forming projection images limited to each layer of retina not only represents information of each layer but also، it would individually display the amount of vulnerability in the each layer caused by a specific disease.

In the first step، a projection image associated with 11 retinal layers was formed with merging the levels and voxels between each pair of boundaries by using different statistical indicators including average، mean، maximum and minimum. Then، retinal layers with more information through statistical indicators were fused with each other to gain an image without any noise and other deficiencies to possess a better clarify and contrast of specific information in X-Y axis.

Using different statistical methods such as average، mean، maximum، minimum and variance، projection images associated with each layer of retina were gotten. Each of these methods made better images in the specific layers than other methods used in the next steps. Fusing of layers with each other was also provided appropriate information from retinal depth in different parts. Contrast enhancement of images related with layers of fused images and creating more complete coordinate of vessels presence were important characteristics of final image.

Resulting projection image would be used more effectively in the extraction of important characteristics of retina including vessels extraction as well as determination of confinement and the center of macular region. Moreover، newer multi-resolution transforms such as curvelet transform would be used to obtain better final fused image.