Cone photoreceptors are highly specialized cells responsible for the origin of vision in the human eye. Their inner segments can be noninvasively visualized using adaptive optics scanning light ophthalmoscopes (AOSLOs) with nonconfocal split detection capabilities. Monitoring the number of cones can lead to more precise metrics for real-time diagnosis and assessment of disease progression. Cell identification in split detection AOSLO images is hindered by cell regions with heterogeneous intensity arising from shadowing effects and low contrast boundaries due to overlying blood vessels. Here, we present a multi-scale circular voting approach to overcome these challenges through the novel combination of: 1) iterative circular voting to identify candidate cells based on their circular structures, 2) a multi-scale strategy to identify the optimal circular voting response, and 3) clustering to improve robustness while removing false positives. We acquired images from three healthy subjects at various locations on the retina and manually labeled cell locations to create ground-truth for evaluating the detection accuracy. The images span a large range of cell densities. The overall recall, precision, and F1 score were 91±4%, 84±10%, and 87±7% (Mean±SD). Results showed that our method for the identification of cone photoreceptor inner segments performs well even with low contrast cell boundaries and vessel obscuration. These encouraging results demonstrate that the proposed approach can robustly and accurately identify cells in split detection AOSLO images.
By the time most retinal diseases are diagnosed, macroscopic irreversible cellular loss has already occurred. Earlier detection of subtle structural changes at the single photoreceptor level is now possible, using the adaptive optics scanning light ophthalmoscope (AOSLO). This work aims to develop a fully automatic segmentation framework to extract cell boundaries from non-confocal split-detection AOSLO images of the cone photoreceptor mosaic in the living human eye. Significant challenges include anisotropy, heterogeneous cell regions arising from shading effects, and low contrast between cells and background. To overcome these challenges, we propose the use of: 1) multi-scale Hessian response to detect heterogeneous cell regions, 2) convex hulls to create boundary templates, and 3) circularlyconstrained geodesic active contours to refine cell boundaries. We acquired images from three healthy subjects at eccentric retinal regions and manually contoured cells to generate ground-truth for evaluating segmentation accuracy. Dice coefficient, relative absolute area difference, and average contour distance were 82±2%, 11±6%, and 2.0±0.2 pixels (Mean±SD), respectively. We find that strong shading effects from vessels are a main factor that causes cell oversegmentation and false segmentation of non-cell regions. Our segmentation algorithm can automatically and accurately segment photoreceptor cells on non-confocal AOSLO images, which is the first step in longitudinal tracking of cellular changes in the individual eye over the time course of disease progression.
The short focal length of the mouse eye gives rise to an optically thick retina (50 D). If in addition, multiple wavelengths
are to be used simultaneously to image an arbitrary combination of retinal layers, the ≈ 10 D of longitudinal chromatic
aberration means a total of 60 D of vergence must be covered. This dictates that marginal rays will cover a wide range of
angles with respect to the optical axis at the pupil of a mouse (or murine) adaptive optics ophthalmoscope, in order to
section through the entire retina with any wavelength simultaneously. In this work, we discuss the compromises
associated with the design of a mouse adaptive optics ophthalmoscope using off-the-shelf spherical reflective and
The incorporation of adaptive optics to scanning ophthalmoscopes (AOSOs) has allowed for in vivo, noninvasive
imaging of the human rod and cone photoreceptor mosaics. Light safety restrictions and power limitations of
the current low-coherence light sources available for imaging result in each individual raw image having a low
signal to noise ratio (SNR). To date, the only approach used to increase the SNR has been to collect large number
of raw images (N >50), to register them to remove the distortions due to involuntary eye motion, and then
to average them. The large amplitude of involuntary eye motion with respect to the AOSO field of view (FOV)
dictates that an even larger number of images need to be collected at each retinal location to ensure adequate
SNR over the feature of interest. Compensating for eye motion during image acquisition to keep the feature of
interest within the FOV could reduce the number of raw frames required per retinal feature, therefore significantly
reduce the imaging time, storage requirements, post-processing times and, more importantly, subject's exposure
to light. In this paper, we present a particular implementation of an AOSO, termed the adaptive optics scanning
light ophthalmoscope (AOSLO) equipped with a simple eye tracking system capable of compensating for eye
drift by estimating the eye motion from the raw frames and by using a tip-tilt mirror to compensate for it in
a closed-loop. Multiple control strategies were evaluated to minimize the image distortion introduced by the
tracker itself. Also, linear, quadratic and Kalman filter motion prediction algorithms were implemented and
tested and tested using both simulated motion (sinusoidal motion with varying frequencies) and human subjects.
The residual displacement of the retinal features was used to compare the performance of the different correction
strategies and prediction methods.
We recently developed several versions of a multimodal adaptive optics (AO) retinal imager, which includes highresolution
scanning laser ophthalmoscopy (SLO) and Fourier domain optical coherence tomography (FDOCT) imaging
channels as well as an auxiliary wide-field line scanning ophthalmoscope (LSO). Some versions have also been
equipped with a fluorescence channel and a retinal tracker. We describe the performance of three key features of the
multimodal AO system including: simultaneous SLO/OCT imaging, which allows SLO/OCT co-registration; a small
animal imaging port, which adjusts the beam diameter at the pupil from 7.5 to 2.5 mm for use with small animals
ubiquitous in biological research or for extended depth-of-focus imaging in humans; and slow scan Doppler flowmetry
imaging using the wide field auxiliary LSO imaging channel. The systems are currently deployed in several
ophthalmology clinics and research laboratories and several investigations have commenced on patients with a variety
of retinal diseases and animals in vision research.
Scanning laser ophthalmoscopes (SLOs) and optical coherence tomographs are the state-of-the-art retinal imaging
instruments, and are essential for early and reliable diagnosis of eye disease. Recently, with the incorporation of adaptive
optics (AO), these instruments have started to deliver near diffraction-limited performance in both humans and animal
models, enabling the resolution of the retinal ganglion cell bodies, their processes, the cone photoreceptor and the retinal
pigment epithelial cells mosaics. Unfortunately, these novel instruments have not delivered consistent performance
across human subjects and animal models. One of the limitations of current instruments is the astigmatism in the pupil
and imaging planes, which degrades image quality, by preventing the wavefront sensor from measuring aberrations with
high spatial content. This astigmatism is introduced by the sequence of off-axis reflective elements, typically spherical
mirrors, used for relaying pupil and imaging planes. Expressions for minimal astigmatism on the image and pupil planes
in off-axis reflective afocal telescopes formed by pairs of spherical mirrors are presented. The formulas, derived from the
marginal ray fans equation, are valid for small angles of incidence (≤15°), and can be used to design laser cavities,
spectrographs and vision adaptive optics systems. An example related to this last application is discussed.
Adaptive optics (AO) is used to correct wavefront aberrations in light in real-time. An AO system is principally made up
of three parts; a wavefront measuring device, a correction device, and a control algorithm to compute the residuals between the measured and a reference wavefront. Deformable mirrors (DM) are commonly used as the correction devices in such a system. This paper presents a method to improve a DM's temporal performance by attenuating parasite oscillations of its reflective membrane when applying high-frequency signals to the mirror actuators. The method consists of implementing low-pass filtering into the software driving the mirror. Different filtering functions were studied both when stimulating one single actuator, and when applying voltages to the complete array of actuators. A linear decomposition in 41 substeps showed the best performance for all considered configurations. The obtained results represented an important
reduction of the settling time as well as the overshoot of the signal response.
This work briefly reviews the achievements of adaptive optics scanning laser ophthalmoscopy to date. Then, an
instrument designed for testing phase imaging modalities is described, and finally, the requirements for MEMS
devices in scanning ophthalmic devices are discussed.
A number of reflective wavefront correctors used in adaptive optics are based on the use of piezoelectric effect, either in piston, tip/tilt or curvature devices. The relation between the voltage applied to drive these devices and the mechanical response always presents hysteresis to some extent. In this work we study the performance of Preisach's classical and non-linear models of hysteresis on a bimorph mirror, which is a curvature device, but both
models can also be applied to piston and tip/tilt devices. Bimorph mirrors with PZT actuators and a passive glass substrate were tested in an adaptive optics test-bed (AOTB) using a Shack-Hartmann wavefront sensor. First- and second-order reversal curves were sampled uniformly in Preisach space, and interpolation algorithms
were implemented to test Preisach's classical and non-linear forward models respectively. Then, arbitrary voltage configuration sequences were applied to the mirror and the responses were recorded. Finally, the inversion of the models was implemented and included in the AOTB linear control algorithm to test the closed-loop performance. We found that both hysteresis models provide a similar improvement in the open-loop error. The improvement estimation depends on the particular sequence applied, the number of samples of the Preisach function and noise among other factors. Finally, we present data showing that the hysteretic behavior in a multi-electrode mirror
is, within experimental error, independent of the electrode geometry, area and location.
We present some of the experimental and theoretical studies in lasers and optics that have currently been performed in our country. We present an overview of several active areas of research that include nonlinear optic spectroscopy, applied optics, and theoretical modeling of laser dynamics.