van den Berg, Thomas J. T. P. PhD
To the Editor: Intra- and Intersession Repeatability of a Double-Pass Instrument
The article by Vilaseca et al.1 focuses on repeatability of the recordings from the OQAS double-pass (DP) instrument. However, of interest also is the question of the validity of its outcome measures for vision, because it uses infrared (IR) radiation (780 nm). Many studies have shown that the results of DP recordings strongly depend on wavelength. As visual sensitivity peaks at 550 nm, the relevance of IR recordings for assessing visual performance is at issue.
Pioneers of DP realized years ago that proper DP recording of the point-spread function (PSF) rests on the question of whether reflectivity of the ocular fundus is well behaved. This was soon found to be a problem. As can be recognized from retinoscopic observation, a significant portion of the light reflected by the fundus stems from the blood-rich choroid and is dependent on pigmentation and wavelength. Charman and Jennings2 noted that the DP-derived modulation transfer function (MTF) for red light deviated strongly from the mid-spectrum one. More than three decades earlier, it was already suspected that red light penetrating into the deeper layers of the fundus such as the choroid, and the subsequent diffuse back scatter, added a background signal to the recording. Indeed, later quantitative studies on the reflective properties of the ocular fundus have substantiated this.3,4 Recently, Logean et al.5 found that even at 543 nm, the problem arose. The half width of the DP-derived PSF was significantly wider for blue-eyed than for dark-eyed individuals.5 To understand these relationships, it is necessary to consider the PSF dynamics.6 The PSF declines by a factor of 1000 in the first 10 min of arc for normal eyes,6 and the rest of the DP recording is dominated by the background. To use a metaphor, the DP-recorded PSF is just similar to a tall broadcasting tower standing on a wide shallow hill, the hill becoming much higher with increasing wavelength of observation. The paracentral dominance of background can clearly be seen in Figs. 6, 7 from ref. 7 showing the difference between 780 and 543 nm DP recordings and in other sources.
A prominent outcome measure of the OQAS is called the “objective scatter index” (OSI). The paracentral part of the PSF between 12 and 20 min of arc (Ip) is taken as representative for scatter in the ocular media. However, the intensity recorded in the 0 to 1 min of arc area (Ic) is also added to the definition, suggested to be OSI = Ip/Ic.1 However, this is not entirely accurate, as the OSI is further divided by 10 (A. Benito at ARVO 2010, personal communication; and confirmed by P. Artal on July 28, 2010, personal communication). With this knowledge, we can compare OSI outcomes to existing literature. From Fig. 2 of Vos et al.,8 it can be seen that for the normal PSF, Ip ≈ 10% and Ic ≈ 35%, so that OSI should be ≈0.029. The current article1 gives a value of 0.32, and another article on the OQAS reports a value of 0.47.9 This approximate 10-fold discrepancy is caused by the high background signal induced by the OQAS using IR light. Using the divisor of 10 may have caused this discrepancy to have gone unnoticed.
Because, with IR radiation, the parameter Ip will not be very sensitive for differences in optical quality, OSI will mainly reflect differences in Ic, the central peak of the PSF. Thus, reported9 variation of OSI with early cataract formation may be understood as a variation in the central peak. Indeed, unpublished data collected with the OQAS in St. Thomas' Hospital in London and the Mayo Clinic in Rochester showed a near-perfect correlation of r = 0.91 and r = 0.92, respectively, between the PSF half-width (50% value) and OSI. This central peak dependence may also explain the high degree of correlation between visual acuity and OSI. Only in low quality ocular media may Ip reflect true para-central PSF. Therefore, coining this outcome measure “Objective Scatter Index” is not correct. Effectively, OSI is a constant divided by the central peak.
Because of space limitations, I present short notes on other aspects. Another outcome parameter of the OQAS, the Strehl ratio, was reported to be 0.271 and 0.23.9 At first glance, this seems to correspond to literature, with reported Strehl ratios in the visible wavelength range around 0.15.10 This surprising outcome can now be understood by the noting that the common definition of the Strehl ratio was not used in the authors' publication, but rather a custom made “Strehl2D.”1 In the absence of a clear literature base, exploratory modeling indicated that the Strehl2D is much larger than the usual Strehl3D. The deviant OQAS definition of Strehl ratio was unknown to authors using this instrument.9 Also, the reported1,9 MTFcutoff values for the OQAS are much lower than those for the visible wavelength range.10 Although aberrometers also use fundus reflectance in IR light, they do not suffer from this problem, because they only use the location of the centroid.
In conclusion, the outcome measures of the OQAS should be interpreted with great caution, because they are obtained with IR radiation, rather than white or 550 nm light. Potentially, the outcome measures relating to the central peak, in particular the 50% value, are sufficiently close to those for visible light to be used. The paracentral part of the PSF is dwarfed by background from the choroid, rendering the OSI parameter largely invalid.
The author thank David Spalton, Mayank Nanavaty, Jay McLaren, and Sanjay Patel for the usage of data.
Thomas J.T.P. van den Berg is an “inventor” to a patent owned by the Royal Academy of The Netherlands and licensed to Oculus for the C-Quant straylight meter.
Thomas J. T. P. van den Berg, PhD
Netherlands Institute for Neuroscience
Amsterdam, The Netherlands
1. Vilaseca M, Peris E, Pujol J, Borras R, Arjona M. Intra- and intersession repeatability of a double-pass instrument. Optom Vis Sci 2010;87:675–81.
2. Charman WN, Jennings JA. The optical quality of the monochromatic retinal image as a function of focus. Br J Physiol Opt 1976;31:119–34.
3. Van Norren D, Tiemeijer LF. Spectral reflectance of the human eye. Vision Res 1986;26:313–20.
4. Delori FC, Pflibsen KP. Spectral reflectance of the human ocular fundus. Appl Opt 1989;28:1061–77.
5. Logean E, Dalimier E, Dainty C. Measured double-pass intensity point-spread function after adaptive optics correction of ocular aberrations. Opt Express 2008;16:17348–57.
6. van den Berg TJ, Franssen L, Coppens JE. Straylight in the human eye: testing objectivity and optical character of the psychophysical measurement. Ophthalmic Physiol Opt 2009;29:345–50.
7. Lopez-Gil N, Artal P. Comparison of double-pass estimates of the retinal-image quality obtained with green and near-infrared light. J Opt Soc Am (A) 1997;14:961–71.
8. Vos JJ, Walraven J, van Meeteren A. Light profiles of the foveal image of a point source. Vision Res 1976;16:215–9.
9. Saad A, Saab M, Gatinel D. Repeatability of measurements with a double-pass system. J Cataract Refract Surg 2010;36:28–33.
10. Navarro R, Artal P, Williams DR. Modulation transfer of the human eye as a function of retinal eccentricity. J Opt Soc Am (A) 1993;10:201–12.