Abstract
Abstract: :
Purpose:To obtain accurate graphic representation of normal foveolas from optimized OCT scans. Methods:OCT scans of classic retinal diseases were digitally modified into five different graphic algorithms: Gaussian smoothing, 32bit Grayscale, Soft Find Edges, Solid Find Edges and Bit Scaling in Corel Photo Paint 10TM software. In order to find the best algorithm, five different retina specialists analysed obtained images and their resolution, comparing details of foveal architecture Subsequently, OCTs of ten normal patients were carried out. Three 4mm horizontal and three 4mm vertical scans of the central fovea were performed. The deepest horizontal and vertical scans were selected. The 1.5mm central curvature (corresponding to the histologic fovea) was determined through high zoom manual drawing applying the high resolution algorithm chosen by the specialists to these scans. Results:Horizontal and vertical curves were obtained corresponding to the central 5 degrees from the umbus. Foveolar topographic representation was achieved connecting horizontal and vertical deepest optimized scans through Bit Scaling pixel modification. Conclusions:This method allows the analysis of graphic representation of a normal foveolar curvature and might help ophthalmologists to correlate vision abnormalities such as hyperacuity and newer aberrations in adaptive optics with corresponding changes of different anatomic patterns.
Keywords: anatomy • image processing • macula/fovea