A scanning laser confocal microscope was used to visualize the human fundus (the back portion of the eyeball, as seen by means of the ophthalmoscope) in vivo from near the retinal surface to deep within the optic nerve head. Thirty-two optical sections based on reflected light were acquired, digitized and aligned to compensate for small eye movements. The registered stack of optical sections was reconstructed with three-dimensional volume rendering software and presented as a movie. The three-dimensional rendering presents a new technique to view the optic nerve. This technique is clinically important since morphological changes of the lamina cribrosa of the optic nerve may be involved in glaucoma.
© Optical Society of America
The morphology (a word coined by Goethe) of the optic nerve from a cow eye was investigated by Leeuwenhoek with his single lens microscope over 300 years ago. Leeuwenhoek was interested in this structure since it could answer the question of how an image could pass from the eye to the brain . Today, researchers are still investigating the morphology of the optic nerve and its relationship to glaucoma.
The human retina contains not only cells that function as light detectors, but also cells involved in signal integration and processing. The human sensory retina contains both cone and rod photoreceptors (phototransduction of light) and also several kinds of interconnecting cells that provide negative feedback to enhance spatial contrast . This type of lateral inhibition was first proposed by Ernst Mach and first shown by H. K. Hartline who received the Nobel prize for his experimentation validation .
The optic disk (1.5 × 2.0 mm) is the site of the eye where all the retinal ganglion cell axons leave the eye via the optic nerve. The lamina cribrosa traverses the width of the optic nerve just posterior to the optic nerve head. Many nerve fiber bundles that comprise the optic nerve pass through a series of pores within the lamina cribrosa.
The retina is a portion of the human visual sensory system that can be observed from outside of the body. However, an external observer cannot just look into the eye and see anything; this magnificent event requires an optical instrument called an ophthalmoscope . The origins of ophthalmoscopy (retinal imaging) reside with both Charles Babbage (1847) and Helmholtz (1851). These scientists invented the first types of ophthalmoscopes that permitted the observer to see the retina inside the living human eye.
One hundred years later Harold Ridley invented electronic ophthalmoscopy and demonstrated the intrinsic advantages of raster scanning a point of light over the retinal surface and electronically detecting the back scattered light to form an image of the retina . Since this work the next major advance was the development by R. Webb of a scanning laser ophthalmoscope which also employs point scanning to illuminate the retina . The scanning laser ophthalmoscope contains a set of apertures of various sizes which permit enhanced resolution due to the principles of confocal microscopy. Confocal microscopy is one of the major recent advances in optical microscopy .
A commercial scanning laser ophthalmoscope was modified by Fitzke and his co-workers for enhanced magnification and improved optical sectioning of the living human retina . This modified instrument was used to acquire, from a human retina in vivo, a stack of optical sections which extended from the optic nerve head through the lamina cribrosa and the optic nerve which connects the eye and the brain.
This paper presents a movie showing the three-dimensional volume rendering of the in vivo human fundus and optic nerve. It represents an interesting example of the use of a light microscope and three-dimensional reconstruction techniques to image a structure of great importance to ophthalmology.
2. Materials and Methods
The details of the techniques have been previously described . A laser scanning ophthalmoscope (Rodenstock model 101, G. Rodenstock Instrumente GmbH, Ottobrunn-Riemerling, Germany) was used to acquire a series of optical sections at different depths near the optic nerve of a living human retina. The light source was a HeNe red laser (633nm red) and the confocal aperture was set to the smallest setting (1 mm, setting number 1) and a field of view of 20 degrees was used. The laser intensity was 200 microwatts. The images were acquired from a normal human subject in the Department of Visual Science, Institute of Ophthalmology, London under the supervision of Dr. F. Fitzke. Informed consent was given by the subject for the imaging of the retina.
Thirty-two confocal microscopic images of the retina and optic nerve were acquired at a rate of 7 frames per second with 8-bit dynamic range. The first optical section was near the surface of the retina at the site of the optic nerve head and the last optical section was approximately 2 mm below the vitreo-retinal interface. The set of images was approximately 4 mm × 5 mm on the retina in the vertical and horizontal directions and was separated by approximately 50 microns in depth.
Because of eye movements due to the cardiac pulse and microsaccades it was necessary to use a cross-correlation technique to align the individual images. This technique has been previously described .
Three-dimensional volume reconstructions of the registered stack of optical sections were performed on a Silicon Graphics, Elan work station with Voxel View software from Vital Images (Fairfield, Iowa). Volume rendering techniques and their applications in three-dimensional reconstructions have been described [9–11]. In order to enhance the three-dimensional visualization the volume rendering is shown as a dynamic movie so that the observer can view not only the retina and optic nerve in en face view but also from a series of rotated orthogonal views.
The movie loop starts with a sequential of the 32 optical sections from near the surface of the retina to deep within the optic nerve head. Near the surface of the retina the branches of the central retinal artery and central retinal vein are shown where they cross the optic nerve head. At the surface of the retina the optic nerve fiber layer is shown as a series of linear structures which converge at the optic nerve head. Posterior to the surface of the optic nerve head the lamina cribrosa is shown. When viewed with the scanning laser ophthalmoscope it appears a bright, highly reflective bean shaped object. The highly reflective region of the lamina cribrosa is surrounded by a darker (less reflective) annulus.
Following the repeated sequence of reflected light confocal microscopic optical sections the dynamic viewing of the three-dimensional optic disk and optic nerve is presented. The three-dimensional reconstruction is rotated on two orthogonal axes to give the viewer a dynamic view of the in vivo human fundus and the optic nerve.
The movie loop shown in this paper represents the first volume rendered three-dimensional reconstruction of the in vivo human optic nerve head, and shows that some of the structures can be tracked through over 1 mm of ocular tissue.
The three-dimensional optical imaging of the optic disk and the optic nerve presents formidable problems since the incident light is rapidly absorbed and scattered as the optical sections penetrate into the deeper regions of the optic nerve . In the fundus that was imaged the melanin pigment absorbs some of the incident light and interferes with imaging some of the deeper layers. This problem may be partially overcome by using longer wavelength infrared illumination. Further investigation of the details of the tissue optics of the retina and the optic nerve may result in improved optical imaging of these important tissue structures.
The clinical optical imaging and the three-dimensional reconstruction of the living human fundus and optic nerve is an important topic in ophthalmology because of alterations in the structure of the lamina cribrosa and their relation to glaucomatous optic nerve damage [14,15].
This work was supported by a grant from NIH EY-06958. The author thanks F. Fitzke, S. Senft, J. Czégé, and the Biomedical Instrumentation Center at USUHS.
References and links
1. B. J. Ford, Single lens, the story of the simple microscope, (Harper & Row, Publishers, New York, 1985), 49–50.
2. J. L. Federman, P. Gouras, H. Schubert, A. M. Slusher, and T. R. Vrabec, Retina and Vitreous, Vol. 9, Textbook of Ophthalmology, S. M. Podos and M. Yanoff, eds. (Mosby-Year Book Europe, Ltd, London, 1994).
4. H. Ridley, “Recent methods of fundus examination including electronic ophthalmoscopy,” Trans. Ophthalmol. Soc. U. K. 72, 497–509 (1952).
5. R. H. Webb, “Scanning laser ophthalmoscope,” in Noninvasive Diagnostic Techniques in Ophthalmology, B. R. Masters, ed. (Springer-Verlag, New York1990), 438–450. [CrossRef]
6. B. R. Masters, Selected Papers on Confocal Microscopy, (SPIE, The International Society for Optical Engineering, Bellingham, WA, 1996).
7. F. W. Fitzke, H. Woon, G. Timberlake, L. Robinson, J. Marshall, and A. C. Bird, “Optical modifications to a scanning laser ophthalmoscope for high magnification, narrow optical section imaging,” Lasers Light Ophthalmol. 4, 7–14 (1991).
9. M. Levoy, “Display of surfaces from Volume Data,” Ph.D. dissertation, (University of North Carolina, Chapel Hill, N.C.1989).
10. N. S. White, “Visualization Systems for Multidimensional CLSM Images,” in Handbook of Biological Confocal Microscopy,J. B. Pawley, ed., (Plenum Press, New York1990), 211–254.
11. A. Kriete, Visualization in Biomedical Microscopies, (VCH, New York1992).
12. O. Minet, G. Müller, and J. Beuthan, Selected Papers on Optical Tomography, Fundamentals and Applications in Medicine, (SPIE, The International Society for Optical Engineering, Bellingham, WA, 1998).
13. V. Tuchin, Selected Papers on Tissue Optics, Applications in Medical Diagnostics and Therapy, (SPIE, The International Society for Optical Engineering, Bellingham, WA, 1994).
14. H. A. Quigley and E. M. Addicks, “Regional differences in the structure of the lamina cribrosa and their relation to glaucomatous optic nerve damage,” Arch. Ophthalmol. 99137–143 (1981). [CrossRef]