| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH |
|
|
|
|||||||
|
|||||||||
Electronic Letters to:
|
|
Electronic letters published:
![[Read eLetter]](/icons/misc/sectionDOWN.gif){kind=icon}
Authors' Response: Optical Coherence Tomography of Retinal Nerve Fiber Layer
Optical Coherence Tomography of Retinal Nerve Fiber Layer
|
|
|||
|
Devinder Chauhan St. Thomas' Hospital, London, UK
Send letter to journal:
chauhands{at}aol.com Devinder Chauhan
|
We should like to thank Drs. Knighton and Huang for their comments on the potential difficulties in imaging the retinal nerve fibre layer (RNFL) with optical coherence tomography (OCT). They have described, both in their letter and previously, the highly directional reflectance of the RNFL and estimate that light incident at just 3.5° off normal will result in a ten-fold reduction in reflectance.1,2 Notwithstanding this, they imply that this is not of significance in vivo, where “the OCT beam is constrained by the pupil to lie near the perpendicular”. In vivo, however, it is often necessary to dilate the pupil, establish the patient’s fixation on one of the instrument’s internal targets, and move the scan line to the region of interest on the retina. This will almost certainly result in the angle of incidence of the OCT light on the retina being both variable and non-perpendicular, leading to errors in the estimation of RNFL of the nature to which the correspondents allude; a lower signal is likely to lead to an underestimation of the RNFL reflectance and, consequently, thickness. Even with this potential experimental error, our most significant finding was that the innermost band of high OCT signal (presumed RNFL) was 7.3 times thicker than the RNFL, not thinner. Thus, if Knighton and Huang are correct in their observations, if we had maintained normal incidence of OCT light on the retina, the signal from this band would be even brighter and thicker still. Knighton and Huang also offer evidence for the existence of an RNFL specific inner OCT signal by referring to three human studies that demonstrate a variation in thickness of this layer “in a manner similar to the known anatomy of the RNFL." The first paper that they quote states that “the location of the outer margin of the RNFL is not precisely defined in the tomographs” and makes no direct comparison with anatomy.3 The second paper quoted4 also fails to make a direct comparison with anatomy and is not in agreement with published normal values.5 The third paper relates OCT findings to visual fields and red-free photography in pathological cases, without reference to published anatomical studies.6 Our study did not wholly refute the notion of a tissue-specific signal. It did demonstrate, however, that the innermost band of high OCT signal is only partly generated by the RNFL. Studies involving control of the imaging geometry, as suggested by Knighton and Huang, may help resolve this broad band of signal into its component parts. Devinder Chauhan References 1. Knighton RW, Baverez C, Bhattacharya A. The directional reflectance
of the retinal nerve fiber layer of the toad. Invest Ophthalmol Vis Sci.
1992;33:2603-2611.
|
|||
|
|
|||
|
Robert Knighton Bascom Palmer Eye Institute, University of Miami School of Medicine, FL
Send letter to journal:
rknighto{at}mednet.med.miami.edu Robert Knighton
|
In a recent paper,1 Chauhan and Marshall interpreted in vitro optical coherence tomography (OCT) images of bovine and human retinas and concluded that an inner band of OCT reflectance was not specific for the retinal nerve fiber layer (RNFL). Their paper does not address a variable that is crucial for in vitro imaging of the RNFL, namely, the angle of incidence of the OCT beam upon the tissue. The reflectance of the RNFL is highly directional;2-4 unless the OCT beam is oriented perpendicular to the nerve fiber bundles, the light reflected by the RNFL will be scattered away from the instrument aperture and the RNFL signal will be diminished or absent. With in vivo imaging of the posterior pole the OCT beam is constrained by the pupil of the eye to lie near the perpendicular. With in vitro imaging it is not so constrained and the angle must be controlled by the experimenter. A quantitative example will illustrate the sensitivity of RNFL reflectance to incident angle. The angular spread of near-infrared light scattering by the rat RNFL has been described by a decaying exponential with a 7° space constant.4 For techniques such as OCT in which the illuminating and collecting apertures move together the effective space constant would be one-half that, or 3.5°. Thus, a deviation from perpendicular of only 8° would cause the RNFL signal to decrease by a factor of ten. The directional reflectance of the RNFL arises from the cylindrical arrangement of nerve fiber bundles.3,4 Because the arrangement and sub- cellular anatomy of unmyelinated retinal axons is similar across vertebrate species, the directional reflectance of the RNFL should be similar also. Indeed, variation in RNFL brightness with viewing angle is well known to clinical ophthalmoscopists.5,6 Furthermore, and relevant to the ablation experiments reported by Chauhan and Marshall,1 glutaraldehyde fixation does not abolish the directional reflectance of rat RNFL (unpublished observations). Considerable evidence suggests that an RNFL specific OCT signal does exist. Studies in humans demonstrate an inner OCT signal that varies in thickness in a manner similar to the known anatomy of the RNFL.7-9 In addition, Huang, et al.,10 correlated quantitative OCT obtained from chicken retina in vivo with histology of the same retinal locations. They showed a high inner OCT peak followed by a lower inner plateau. This peak was wider in retinal areas where the RNFL was thicker, implying a correspondence between the two. Future quantitative comparisons of OCT and histology will undoubtedly increase our understanding of the RNFL signal, but we suggest that to be most useful these comparisons must include precise control of the imaging geometry. Robert W. Knighton References 1. Chauhan DS, Marshall J. The interpretation of optical coherence
tomography images of the retina. Invest Ophthalmol Vis Sci. 1999;40:2332-
2342.
|
|||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH |
eLetters: