We read with interest the article by Neitz and Neitz
1 titled “Diffusion Optics Technology (DOT): A Myopia Control Spectacle Lens Based on Contrast Theory.” Based on the finding that mutations that cause the longer wavelength cones to randomly be either normal or of impaired efficiency are associated with high myopia, they proposed that this was because the heterogeneous cone mosaic signaled a spuriously high contrast. From this insight, they developed a novel myopia control lens that reduces image contrast and which has shown significant clinical effectiveness. Although other myopia control lenses typically attempt to create a peripheral myopic defocus, a recent optical analysis of some of these lenses suggests that they mostly do not, and in fact they may simply be reducing peripheral image contrast as are the DOT lenses.
2 The findings of Neitz and Neitz, and their collaborators have been some of the most exciting and important in the field of myopia research in some time.
However, we would like to propose another explanation for these results. It is an oversimplification to say that current anti-myopia lenses reduce image contrast; rather, they reduce image contrast at the higher spatial frequencies and preserve contrast at the lower spatial frequencies.
3,4 A recent ARVO abstract suggests that the DOT lenses also spare the lower spatial frequencies.
5 In other words, these lenses function as spatial lowpass filters. The evidence from animal models is that emmetropization uses the midrange of spatial frequencies.
6 This should not be surprising, as emmetropization must operate robustly when images are out of focus and higher spatial frequencies are not available. We have argued on theoretical grounds that human emmetropization might rely primarily on image data in the range of (very roughly) 3 to 4 cycles per degree.
7 If these lenses relatively preserve the spatial frequencies that emmetropization uses and filter out spatial frequencies higher than those, then the function would be not to reduce contrast but rather to reduce interfering higher spatial frequency noise.
We illustrate this point with a toy example. We took a standard machine vision camera and fitted it with a 16-mm focal length lens at an aperture of
f/2.8. This obviously does not reproduce the chromatic and other aberrations of the human eye, but it approximates the power of a human eye with a 6-mm pupil. We took pictures of a scene of black-and-white grayscale trees either with the bare lens (panel A of the
Fig.) or through the relatively homogeneous peripheral optics of a MyoCare 0-diopter (plano) lens (Carl Zeiss Vision International, Aalen, Germany) placed in front (panel B of the
Fig.). Panel C of the
Figure illustrates the grayscale pixel intensities of the recorded image taken from a horizontal row of pixels passing through the center of the images. The lens greatly reduced the high spatial frequencies, but the amplitude range of the lower frequencies was largely preserved. Panel D of the
Figure illustrates the two-dimensional radially averaged Fourier amplitude of the entire image: At the lower spatial frequencies, the amplitude is preserved, but the lens progressively cuts this off at higher spatial frequencies.
Comparing the spatial frequency filtering properties of the different myopia control lenses with their clinical efficiency could be of interest and might provide another approach to optimizing their effectiveness.
Supported by a grant from the National Eye Institute, National Institutes of Health (R01EY028578 to TJG); by a UAB Core Grant for Vision Research (P30EY003039) and (R21EY036536 to SK).
Disclosure: T.J. Gawne, optical methods of myopia control (P); S. Khanal, None; T.T. Norton, optical methods of myopia control (P)