Final Report Summary - OPAL (Optical and adaptational limits of vision)
The ITN OpAL combined research efforts around the question how our vision could become so well adapted to the daily needs. The major questions, tackled in clusters, were (1) how neural adaptation can result in subjective image sharpening, (2) how the visual system manages to hide the effects of chromatic aberration, (3) how it deals with defocused images (which often go unnoticed), (4) how it deals with imperfections of the optics at low luminances, (5) whether visual performance may vary over the diurnal cycle, and (6) why light scatter in the ocular media is usually not seen. Considerable progress was made towards the answers of these questions as described below.
It was studied how ocular parameters determine the individual differences in transverse chromatic aberration. Surprisingly, lens thickness and vertical lens tilt could explain a large part of it - somewhat unexpected and it also shows that we are struggeling with more and more transverse chromatic aberration when we get older. The foveal blue scotoma was studied: When human subjects look at a blue screen which in modulated in brightness, they can see their fovea as a black spot. Since most human subjects have no S-cones in the fovea, it is assumed that the black spot reflects the “foveal blue scotoma”. Interestingly, its shape varies considerably from subject to subject but it was not known which factor determine this variability. It was found that it is not due to macular pigment but must rather be due to the lack of S-cones in the foveal center, as initially hypothesized. Interestingly, the diameter of the foveal blue scotoma was found to increase with the steepness of the foveal pit. Related to fundal reflectivity, the calibration of a widely used technique to measure refractive state of an eye from a distance, eccentric photorefraction, is highly variable, but it is not known how and why. It was studied how ocular optical parameters, like axial length, pupil size, fundal reflectance affect the calibration. A fundamental result was that the average pixel brightness of the photoretinoscopic reflex in the pupil is correlated with the conversion factor in the calibration. This project has contributed to the understanding of eccentric photorefraction and is of high interest to all researchers using photorefraction.
Another problem tackled in the OpAL network was to determine the effects of changes in pupil size on the output of a custom-built video eye tracker. Furthermore, the device was then used to describe the impact of cone or rod vision on the patterns of fixational eye movements; in addition, the effect of defocus in the observed image on fixational eye movements was studied. A setup was developed to explore the effects of low order aberrations on the absolute threshold of vision. However, there was no clear evidence that correction of low order aberrations affected the absolute threshold in a statistically significant way. Further tests revealed that in the periphery, where the pooling of the photoreceptors is high, the effect of aberrations was minor, masked by the systematic errors of the process.
Further efforts were directed towards the question how imaging of the ocular fundus is affected by scattering and whether there is an effective way of compensating intraocular scattering during fundus imaging. An existing double pass system was modified to study the spatial properties of fundus reflectance and red-green relative spectral sensitivity psychophysically. It showed that diffused fundus light affects spectral sensitivity of the subjects although the input was small. An optical system was built to measure intraocular scattering for several wavelengths in the visible range, which could also take fundus images at those wavelengths. An algorithm was developed for the compensation of the measured scattering from the fundus images. The method was applied for the calculation of the Macular Pigment Optical Density (MPOD) and showed that intraocular scattering can lead to serious underestimation of MPOD measurements. For the particular subjects the underestimation was between 60 and 70%.
A more subject-friendly version of a double-pass based technique was built to measure intraocular scatter to make the technique available to more users. A new psychophysical method was developed to measure the Contrast Sensitivity Function (CSF). The accuracy of quantification of intraocular scattering of the newly developed method was compared to other, commercially available, methods. Furthermore, different methods for assessing the amount of intraocular scattering were compared, revealing the efficiency and the deficiency of each method. Finally, in a joint effort with network partners, the effect of increased intraocular scattering on the accommodation process was studied. New psychophysical routines were developed to evaluate on- and off-axis vision, in terms of both contrast sensitivity and resolution and detection acuity. The ability of these visual functions to adapt to optical blur were investigated over the visual field. These studies merged in the interesting and unexpected finding that foveal vision was improved after prolonged exposure to defocus blur but that this effect was suppressed when also the periphery was blurred. The result emphasize the importance of also considering the peripheral vision and has important applications e.g. when designing optics that manipulates the image quality on the peripheral retina to reduce the progression of myopia. Measurements with calculated instead of real optical blur were added and provided more information on which features of the image the adaptation occurs. Interestingly, the preliminary results show a similar reduction of the adaptational effect when the stimuli is larger - it appears that local adaptation is more efficient than global, an unexpected and new finding. Furthermore, the effect of astigmatism has been investigated over the visual field. Off-axis detection and resolution acuity have been compared under optimum optical correction with adaptive optics and found to differ as they involve different parts of the neural processing system. Generally, detection acuity via aliasing is better than resolution and resolution acuity for details oriented radially in the visual field, i.e. pointing towards the fovea, is better than other orientations. The latter is known as the meridional effect and was found to be more pronounced for resolution than for detection, as detection is less orientation specific. No evidence was found that the preference to radial gratings is an adaptation to the natural off-axis astigmatism of the eye.
Psychophysical routines were also implemented to evaluate on- and off-axis vision, in terms of both contrast sensitivity and resolution and detection acuity. These visual functions have been measured under optimum monochromatic optical correction with adaptive optics at the same time as different amounts of transverse chromatic aberrations (TCA) was induced by trial lenses. The study merged into the important finding that the effects of spectacle induced TCA affect peripheral detection acuity, maybe even to a larger degree than in central vision. It is therefore important to avoid inducing additional peripheral TCA, e.g. in spectacles for people with central visual field loss. The measurement protocol with induced TCA can also be used as a subjective method to measure TCA in human subjects on- and off-axis. However, it is quite time-consuming and demanding on the subject and the project was therefore extended to aim for objective methods. A collaboration project was established with Prof. Austin Roorda at the University of California, Berkeley, and the first ever objective measurements of the ocular TCA was performed in their adaptive optics scanning laser ophthalmoscope. The change of TCA over the central 30° field was similar to that predicted by theory, but the absolute values varied between subjects. Interestingly, no improvement of visual function was found in quasi monochromatic light of 543.5 nm after adaptive optics correction of monochromatic aberration, neither in peripheral contrast sensitivity nor in detection acuity. The lack of improvements might be due to the overall reduction in luminance and further investigations are needed.
In another study, adaptation to blur was explored as produced by a special pattern, that achieved bifocal presbyopic corrections, and found that subjects are indeed capable of adaptation to blur produced by simultaneous vision. Adaptations of the eye to its own aberrations were also studied. Previous studies had shown that the eye is in fact adapted to its native blur magnitude and orientation. Novel results show now this effect from a binocular perspective, and found that the visual system is in fact calibrated to the magnitude and orientation and blur of the eye with least optical aberrations. These results have important implications to yet another type of presbyopic correction (monovision). Studies on the adaptation to blur focused on a particular type of blur (bifocal blur), of clinical relevance, due to the increasing use of multifocal corrections of presbyopia. These studies involved the development of a new system, a simultaneous vision simulator. The study of neural adaptation to native blur has been addressed binocularly, which represents an expansion over initially planned experiments on one eye only. These studies were conducted using a custom Adaptive Optics system.
Studies were undertaken to better understand the sources of aberrations in the crystalline lens (natural, accommodated and artificial). By using custom-developed Optical Coherence Tomography and wavefront aberrometry as well as customized model eyes, the understanding on the contribution of the lens aberrations into the eye optics was deepened. The results have an impact in the understanding of the lens, accommodation, and intraocular lens design. The project has investigated lens surface topography (radii of curvature, astigmatism and high order aberration surface elevations) in fully accommodated isolated human lenses, as well as a new method (OCT-based) for measuring tilt and decentration of lenses in vivo. The project has produced the first OCT fully based eye models, which have allowed understanding of the contributing factors to ocular optical degradation.
Pilot measurements were done to simulate visual acuity in presence of different optical aberrations in a limited number of subjects. The framework for data analysis and approximation using Artificial Neural Networks was developed and demonstrated. The reversed approach, inducing optical imperfections and modeling visual performance was continued. An adaptive optics system was developed to simulate predetermined aberrations and scattering in the human eye and to generate phase maps corresponding to aberrations and scatter associated to ageing in the human eye. A new integrated optical instrument was built for measurements of straylight into a portable optoelectronic device suitable for measurements in clinical environments. Two parallel studies were organised in the University Hospital of Heraklion and the University Hospital of Murcia. In these studies, subjects with cataract were evaluated with the instrument and additionally their visual performance was evaluated using standard psychophysical tests. The results in both studies indicated that the intraocular scattering measured with the new optical instrument correlated well with psychophysical data and clinical grading of cataract densities. Additionally, diffusers were defined in light scattering studies to normalize the calibration of instruments.
OpAL’s private sector partner developed various lenses that were provided to the partners. Very careful and extensive studies were performed to detect small variations in refractive state over the day. These studies showed that there are indeed variations with most myopic refractions around 3 p.m. but no corresponding change in eye length, measured with low coherence interferometry. It was also studied how exposure to positive defocus affect short-term ocular biometry. Already after 30 minutes the length of the eye had increased significantly. In the course of an intense collaboration between academic and private partners, adaptation to TCA was examined. This goal required a technique to measure TCA subjectively. A Vernier alignment method was chosen. While trying to characterize the impact of TCA on vision, it became clear that LCD screens could not render the effects of TCA correctly. Thus, a LED system was developed. The central and peripheral visual acuity was measured with different amounts of imposed TCA. Results were that peripheral vision was more affected by equal amounts of induced TCA than central vision. The question whether the TCA-detection is dependent from color spectra could not yet be fully clarified. Adaptation to induced TCA appeared slightly variable between subjects. However, most subjects perceived less TCA after wearing the TCA-inducing lenses for short periods of time.
In summary, the OpAL network trained a considerable number of students in the field, generated many meeting presentations and original research articles (a major number still to be published), and provided the fellows with the opportunity to build networks with the other fellows and in particular with established scientists in the field.
WEBSITE: www.itn-opal.eu
________________________________________________________________
RESEARCH COORDINATOR: Prof. Dr. Frank Schaeffel, Eberhard Karls Universitaet Tuebingen,
frank.schaeffel@uni-tuebingen.de
MANAGEMENT: Dr. Thomas Wheeler-Schilling; Dr. Michaela Bitzer, Eberhard Karls Universitaet
Tuebingen; thomas.wheeler-schilling@uni-tuebingen.de; michaela.bitzer@klinikum.uni-tuebingen.de
SEE ALSO ANNEX I
It was studied how ocular parameters determine the individual differences in transverse chromatic aberration. Surprisingly, lens thickness and vertical lens tilt could explain a large part of it - somewhat unexpected and it also shows that we are struggeling with more and more transverse chromatic aberration when we get older. The foveal blue scotoma was studied: When human subjects look at a blue screen which in modulated in brightness, they can see their fovea as a black spot. Since most human subjects have no S-cones in the fovea, it is assumed that the black spot reflects the “foveal blue scotoma”. Interestingly, its shape varies considerably from subject to subject but it was not known which factor determine this variability. It was found that it is not due to macular pigment but must rather be due to the lack of S-cones in the foveal center, as initially hypothesized. Interestingly, the diameter of the foveal blue scotoma was found to increase with the steepness of the foveal pit. Related to fundal reflectivity, the calibration of a widely used technique to measure refractive state of an eye from a distance, eccentric photorefraction, is highly variable, but it is not known how and why. It was studied how ocular optical parameters, like axial length, pupil size, fundal reflectance affect the calibration. A fundamental result was that the average pixel brightness of the photoretinoscopic reflex in the pupil is correlated with the conversion factor in the calibration. This project has contributed to the understanding of eccentric photorefraction and is of high interest to all researchers using photorefraction.
Another problem tackled in the OpAL network was to determine the effects of changes in pupil size on the output of a custom-built video eye tracker. Furthermore, the device was then used to describe the impact of cone or rod vision on the patterns of fixational eye movements; in addition, the effect of defocus in the observed image on fixational eye movements was studied. A setup was developed to explore the effects of low order aberrations on the absolute threshold of vision. However, there was no clear evidence that correction of low order aberrations affected the absolute threshold in a statistically significant way. Further tests revealed that in the periphery, where the pooling of the photoreceptors is high, the effect of aberrations was minor, masked by the systematic errors of the process.
Further efforts were directed towards the question how imaging of the ocular fundus is affected by scattering and whether there is an effective way of compensating intraocular scattering during fundus imaging. An existing double pass system was modified to study the spatial properties of fundus reflectance and red-green relative spectral sensitivity psychophysically. It showed that diffused fundus light affects spectral sensitivity of the subjects although the input was small. An optical system was built to measure intraocular scattering for several wavelengths in the visible range, which could also take fundus images at those wavelengths. An algorithm was developed for the compensation of the measured scattering from the fundus images. The method was applied for the calculation of the Macular Pigment Optical Density (MPOD) and showed that intraocular scattering can lead to serious underestimation of MPOD measurements. For the particular subjects the underestimation was between 60 and 70%.
A more subject-friendly version of a double-pass based technique was built to measure intraocular scatter to make the technique available to more users. A new psychophysical method was developed to measure the Contrast Sensitivity Function (CSF). The accuracy of quantification of intraocular scattering of the newly developed method was compared to other, commercially available, methods. Furthermore, different methods for assessing the amount of intraocular scattering were compared, revealing the efficiency and the deficiency of each method. Finally, in a joint effort with network partners, the effect of increased intraocular scattering on the accommodation process was studied. New psychophysical routines were developed to evaluate on- and off-axis vision, in terms of both contrast sensitivity and resolution and detection acuity. The ability of these visual functions to adapt to optical blur were investigated over the visual field. These studies merged in the interesting and unexpected finding that foveal vision was improved after prolonged exposure to defocus blur but that this effect was suppressed when also the periphery was blurred. The result emphasize the importance of also considering the peripheral vision and has important applications e.g. when designing optics that manipulates the image quality on the peripheral retina to reduce the progression of myopia. Measurements with calculated instead of real optical blur were added and provided more information on which features of the image the adaptation occurs. Interestingly, the preliminary results show a similar reduction of the adaptational effect when the stimuli is larger - it appears that local adaptation is more efficient than global, an unexpected and new finding. Furthermore, the effect of astigmatism has been investigated over the visual field. Off-axis detection and resolution acuity have been compared under optimum optical correction with adaptive optics and found to differ as they involve different parts of the neural processing system. Generally, detection acuity via aliasing is better than resolution and resolution acuity for details oriented radially in the visual field, i.e. pointing towards the fovea, is better than other orientations. The latter is known as the meridional effect and was found to be more pronounced for resolution than for detection, as detection is less orientation specific. No evidence was found that the preference to radial gratings is an adaptation to the natural off-axis astigmatism of the eye.
Psychophysical routines were also implemented to evaluate on- and off-axis vision, in terms of both contrast sensitivity and resolution and detection acuity. These visual functions have been measured under optimum monochromatic optical correction with adaptive optics at the same time as different amounts of transverse chromatic aberrations (TCA) was induced by trial lenses. The study merged into the important finding that the effects of spectacle induced TCA affect peripheral detection acuity, maybe even to a larger degree than in central vision. It is therefore important to avoid inducing additional peripheral TCA, e.g. in spectacles for people with central visual field loss. The measurement protocol with induced TCA can also be used as a subjective method to measure TCA in human subjects on- and off-axis. However, it is quite time-consuming and demanding on the subject and the project was therefore extended to aim for objective methods. A collaboration project was established with Prof. Austin Roorda at the University of California, Berkeley, and the first ever objective measurements of the ocular TCA was performed in their adaptive optics scanning laser ophthalmoscope. The change of TCA over the central 30° field was similar to that predicted by theory, but the absolute values varied between subjects. Interestingly, no improvement of visual function was found in quasi monochromatic light of 543.5 nm after adaptive optics correction of monochromatic aberration, neither in peripheral contrast sensitivity nor in detection acuity. The lack of improvements might be due to the overall reduction in luminance and further investigations are needed.
In another study, adaptation to blur was explored as produced by a special pattern, that achieved bifocal presbyopic corrections, and found that subjects are indeed capable of adaptation to blur produced by simultaneous vision. Adaptations of the eye to its own aberrations were also studied. Previous studies had shown that the eye is in fact adapted to its native blur magnitude and orientation. Novel results show now this effect from a binocular perspective, and found that the visual system is in fact calibrated to the magnitude and orientation and blur of the eye with least optical aberrations. These results have important implications to yet another type of presbyopic correction (monovision). Studies on the adaptation to blur focused on a particular type of blur (bifocal blur), of clinical relevance, due to the increasing use of multifocal corrections of presbyopia. These studies involved the development of a new system, a simultaneous vision simulator. The study of neural adaptation to native blur has been addressed binocularly, which represents an expansion over initially planned experiments on one eye only. These studies were conducted using a custom Adaptive Optics system.
Studies were undertaken to better understand the sources of aberrations in the crystalline lens (natural, accommodated and artificial). By using custom-developed Optical Coherence Tomography and wavefront aberrometry as well as customized model eyes, the understanding on the contribution of the lens aberrations into the eye optics was deepened. The results have an impact in the understanding of the lens, accommodation, and intraocular lens design. The project has investigated lens surface topography (radii of curvature, astigmatism and high order aberration surface elevations) in fully accommodated isolated human lenses, as well as a new method (OCT-based) for measuring tilt and decentration of lenses in vivo. The project has produced the first OCT fully based eye models, which have allowed understanding of the contributing factors to ocular optical degradation.
Pilot measurements were done to simulate visual acuity in presence of different optical aberrations in a limited number of subjects. The framework for data analysis and approximation using Artificial Neural Networks was developed and demonstrated. The reversed approach, inducing optical imperfections and modeling visual performance was continued. An adaptive optics system was developed to simulate predetermined aberrations and scattering in the human eye and to generate phase maps corresponding to aberrations and scatter associated to ageing in the human eye. A new integrated optical instrument was built for measurements of straylight into a portable optoelectronic device suitable for measurements in clinical environments. Two parallel studies were organised in the University Hospital of Heraklion and the University Hospital of Murcia. In these studies, subjects with cataract were evaluated with the instrument and additionally their visual performance was evaluated using standard psychophysical tests. The results in both studies indicated that the intraocular scattering measured with the new optical instrument correlated well with psychophysical data and clinical grading of cataract densities. Additionally, diffusers were defined in light scattering studies to normalize the calibration of instruments.
OpAL’s private sector partner developed various lenses that were provided to the partners. Very careful and extensive studies were performed to detect small variations in refractive state over the day. These studies showed that there are indeed variations with most myopic refractions around 3 p.m. but no corresponding change in eye length, measured with low coherence interferometry. It was also studied how exposure to positive defocus affect short-term ocular biometry. Already after 30 minutes the length of the eye had increased significantly. In the course of an intense collaboration between academic and private partners, adaptation to TCA was examined. This goal required a technique to measure TCA subjectively. A Vernier alignment method was chosen. While trying to characterize the impact of TCA on vision, it became clear that LCD screens could not render the effects of TCA correctly. Thus, a LED system was developed. The central and peripheral visual acuity was measured with different amounts of imposed TCA. Results were that peripheral vision was more affected by equal amounts of induced TCA than central vision. The question whether the TCA-detection is dependent from color spectra could not yet be fully clarified. Adaptation to induced TCA appeared slightly variable between subjects. However, most subjects perceived less TCA after wearing the TCA-inducing lenses for short periods of time.
In summary, the OpAL network trained a considerable number of students in the field, generated many meeting presentations and original research articles (a major number still to be published), and provided the fellows with the opportunity to build networks with the other fellows and in particular with established scientists in the field.
WEBSITE: www.itn-opal.eu
________________________________________________________________
RESEARCH COORDINATOR: Prof. Dr. Frank Schaeffel, Eberhard Karls Universitaet Tuebingen,
frank.schaeffel@uni-tuebingen.de
MANAGEMENT: Dr. Thomas Wheeler-Schilling; Dr. Michaela Bitzer, Eberhard Karls Universitaet
Tuebingen; thomas.wheeler-schilling@uni-tuebingen.de; michaela.bitzer@klinikum.uni-tuebingen.de
SEE ALSO ANNEX I