In this work, we characterize the in vivo ocular optics of the human eye with a lateral resolution of 8.6 microns, which implies roughly 1 million measurement points for a pupil diameter of 9 mm.
Knowledge of the optics of the human eye has allowed us to broaden the understanding of the mechanism of human vision. It has led the current field of ophthalmology to a highly technical level where customized treatments and instrumentation-based diagnostics have become the norm. However, measuring the optics of the eye is not a trivial task. In 1961, Smirnov developed an advanced version of the Scheiner disk1 to subjectively estimate the optical imperfections of the eye, also called aberrations or phase map, in a process named ocular aberrometry2. At that time, he anticipated that his invention would have no practical application, stating that “…the calculations take 10–12 h […] Therefore, it is unlikely that such detailed measurements will ever be adopted by practitioner-ophthalmologists.” He did not foresee that advances in computer science would greatly speed up calculations and that scientists would find the measurement of ocular aberrations an invaluable tool, progressing from a research instrument to a clinical application system. Indeed, some of modern ophthalmology could not be understood without the deep understanding of ocular optics that aberrometers have made possible.
Currently, there are several options for the estimation of ocular aberrations based on different techniques, i.e., Hartmann-Shack sensors (H–S)3, pyramidal sensors (P-S)4, interferometric techniques5, laser ray tracing6 and curvature sensors7. Techniques based on H–S sensors are the most widely used in ophthalmology. However, they are restricted to sampling the phase map at up to 2,600 measurement points within the pupil of the eye (approximately 175 µm of lateral resolution for a 9 mm pupil diameter8) and also suffer from a limited dynamic range9. These limitations constrain its usefulness in abnormal eyes, which are precisely the most interesting to characterize10. Currently, P-S offers the highest resolution in ocular aberrometry, up to 45,000 measurement points (approximately 37 µm theoretical lateral resolution for a 9 mm pupil diameter)8. Nevertheless, P-S generally suffers from non-linear behaviour, diffraction effects between its different pupil images, and a relatively limited dynamic range as a result of the trade-off between the slope accuracy and achievable spatial resolution11. These drawbacks cause the phase maps obtained to be somewhat fuzzy, and the details that can be visualized do not match the theoretical resolution8. Therefore, the H–S sensor is still considered the gold standard in ophthalmology.
Actual knowledge about the optics of the normal human eye can be roughly summarized as follows: it has relatively smooth optics with no optical patterns that diverge too greatly from a conical shape with some astigmatism. Its phase map is usually expressed as a decomposition of orthonormal polynomials within a circular pupil, in particular Zernike polynomials12, where most of the deformations can be described with the first two to three dozen terms13,14. Thus, up to 66 Zernike terms are usually considered sufficient to represent ocular optics with high quality even in pathological eyes, assuming a loss of high frequencies that is not considered to hide any relevant information15,16. However, some authors have already drawn attention to the inefficiency of this approach17.
In this work, we characterize the ocular optics of living human eyes with a lateral resolution of approximately 8.55 µm. This implies more than a million points of measurement for a pupil with a diameter of 9 mm, orders of magnitude higher than the current gold standard offered by the H–S sensor. We show evidence that when ocular optics are measured at a sufficiently high resolution, a series of phase patterns emerge. We hypothesize that this finding could have a great impact on some current ophthalmic surgical procedures and on the clarification of some fundamental mechanisms of human vision.