Adaptation of a support vector machine algorithm for segmentation and visualization of retinal structures in volumetric optical coherence tomography data sets
Recent developments in Fourier domain—optical coherence tomography (Fd-OCT) have increased the acquisition speed of current ophthalmic Fd-OCT instruments sufficiently to allow the acquisition of...
Enhancing the signal-to-noise ratio in ophthalmic optical coherence tomography by image registration—method and clinical examples
Optical coherence tomography (OCT) has already proven an important clinical tool for imaging and diagnosing retinal diseases. Concerning the standard commercial ophthalmic OCT systems, speckle noise i...
Bartosz Sikorski Nicolaus Copernicus University, Collegium Medicum, Department of Ophthalmology, Curie-Sklodowskiej 9, PL-85-094 Bydgoszcz, Poland
Tomasz Bajraszewski Nicolaus Copernicus University, Institute of Physics, ul. Grudziądzka 5/7, PL-87-100 Toruń, Poland
Vivek J. Srinivasan Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, Research Laboratory of Electronics, Cambridge, Massachusetts 02139
Anna Szkulmowska Nicolaus Copernicus University, Institute of Physics, ul. Grudziądzka 5/7, PL-87-100 Toruń, Poland
Jakub J. Kauny Nicolaus Copernicus University, Collegium Medicum, Department of Ophthalmology, Curie-Sklodowskiej 9, PL-85-094 Bydgoszcz, Poland
James G. Fujimoto Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, Research Laboratory of Electronics, Cambridge, Massachusetts 02139
Andrzej Kowalczyk Nicolaus Copernicus University, Institute of Physics, ul. Grudziądzka 5/7, PL-87-100 Toruń, Poland
Wepresent a computationally efficient, semiautomated method for analysis of posteriorretinal layers in three-dimensional (3-D) images obtained by spectral opticalcoherence tomography (SOCT). The method consists of two steps: segmentationof posterior retinal layers and analysis of their thickness anddistance from an outer retinal contour (ORC), which is introducedto approximate the normal position of external interface of thehealthy retinal pigment epithelium (RPE). The algorithm is shown toeffectively segment posterior retina by classifying every pixel in theSOCT tomogram using the similarity of its surroundings to areference set of model pixels from user-selected area(s). Operator interventionis required to assess the quality of segmentation. Thickness anddistance maps from the segmented layers and their analysis arepresented for healthy and pathological retinas.
You are not logged in to this journal. 
u
ny
Connotea
CiteULike
del.icio.us
BibSonomy
