Purchase this article with an account.
I. Gorczynska, V. J. Srinivasan, J. J. Liu, R. W. S. Chen, L. N. Vuong, M. Wojtkowski, J. S. Duker, J. S. Schuman, J. G. Fujimoto; Projection OCT Fundus Imaging for Visualization of 3D-OCT Data. Invest. Ophthalmol. Vis. Sci. 2008;49(13):1887. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
High-speed Fourier/spectral domain OCT enables three dimensional (3D) imaging of retinal pathology. However, analysis of 3D-OCT datasets is challenging. There is a need for methods enhancing visualization of pathology and reducing the need to review large numbers of images. One of the approaches is automatic segmentation of pathology. However, segmentation can be unreliable if the continuity of layers is disrupted or the contrast in images is changed by disease. We developed a method enhancing visualization of retinal structure by displaying pathology in selected retinal levels. This method, projection OCT fundus imaging, requires the detection of only one boundary between retinal layers. Information from levels parallel to this reference plane is displayed in the form of projection fundus images which can be correlated with standard ophthalmic diagnostics.
3-D OCT data sets were acquired using a prototype spectral/Fourier domain OCT instrument. The imaging speed was 24,000 lines/s enabling acquisition of 3D data consisting of 500x180x1024 voxels in ~4 s. Imaging was performed on patients with a cross-section of pathologies. An automatic edge detection algorithm was applied to detect the outer retinal contour. Projection OCT fundus images were created by axial summation of data within levels parallel to the reference surface.
Projection OCT fundus images of retinal pathologies (AMD, central serous chorioretinopathy, vitelliform dystrophy and others) will be presented and compared with standard diagnostics. Projection OCT fundus imaging enhances visualization of pathology by en face filtering of 3D data which rejects information from unwanted retinal layers. This method has the advantage of extracting information relevant for diagnosis without the need to segment all layers affected by the disease. Projection OCT fundus imaging visualizes pathology in a small set of en face images. Images can be color-coded and combined to display the depth position in color. Additionally, en face segmentation in projection images can be performed to enable quantitative measurement of pathology (e.g. area of drusen, RPE atrophy).
Projection OCT fundus imaging is a robust qualitative method of 3D-OCT data visualization which provides information complementary to standard ophthalmic diagnostics.
This PDF is available to Subscribers Only