Purchase this article with an account.
J. Rosen, R. Margolis, S. Schneider, R.C. Gentile, S. McCormick, R. Rosen; Clinical–pathologic Modeling of Optical Coherence Tomography (OCT) Ophthalmoscopy Using Coronal Sectioning of Retinal Specimens, 3–Dimensional Histologic Reconstructions and Computer Simulation of Retinal Scanning . Invest. Ophthalmol. Vis. Sci. 2004;45(13):2379.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: The OCT ophthalmoscope (OCT/SLO) precisely correlates retinal surface detail with internal anatomy improving upon standard B–scan OCT. It does this by employing C–scan OCT images which clinicians are unfamiliar with. The purpose of our study was to develop histologic and computer models to aid in interpretation of C–scan OCT images. Methods: Three strategies were employed to model C–scan OCT images. The first two involved histologic examination of the posterior segments of eye bank globes. OCT Ophthalmoscopic imaging was attempted in selected samples. The retinas were then either embedded in epon resin or paraffin according to standard protocol. Epon embedded samples were sliced transversely in one micron thick sections and stained with toluidine blue to simulate C–scan imaging. Paraffin embedded samples were cut longitudinally in 6–micron thick sections and stained with hematoxylin and eosin. Digital photographs of the sections were reconstructed into interactive 3–D images which could be re–sliced coronally to approximate C–scanning. The third model, an interactive computer simulation of the encounterof the macula with the laser scanning plane was developed utilizing Macromedia software. Results: Transversely cut histologic sections demonstrated how slight tilting of the slicing or scanning axis away from the normal produced various perspective artifacts. These include splaying of retinal layers, patchy appearance of multiple depth structures within individual sections, and variable distortion of size of adjacent structures. The computer simulation was helpful since it allowed unlimited interactive encounters between the scanning plane and the model eye. The main limitation of the animation was lack of detail of microscopic anatomy. The reconstructed blocks of longitudinally cut histologic specimens allowed good appreciation of the microanatomic structures but were limited by fixation artifacts. OCT imaging of the eye cup specimens required multiple averaging and demonstrated artifacts of positioning and fixation. Conclusions: C–scan OCT images can be simulated by a variety of histologic and computer animation models. Each of these models provided different insights into the interaction between the fundus and the scanning plane. The combination of approaches may prove to be potentially useful in teaching interpretation of clinical images and enhancing the diagnostic utility of the instrument.
This PDF is available to Subscribers Only