2.1.

Optical Design: BRBS

The concept of BRBS has been described previously.³ Briefly, a spot of circularly polarized 830-nm laser light is scanned as an annulus on both retinas. For each eye, if the annulus surrounds the center of the fovea, the differential polarization signal of the returning light has a frequency twice the input scanning frequency. If the annulus does not surround the fovea, the differential polarization signal of the reflected light has a primary frequency equal to the input scanning frequency. For the PVS, the scanning frequency is 100 Hz, so that a predominantly 200-Hz signal from an eye indicates central fixation, while a 100-Hz signal indicates paracentral fixation. Thus, analysis of the frequency of the signals representing the right and left eyes can indicate whether a subject is fixating on the target with one eye, both eyes, or neither eye.

Figure 1 details the light pathway for the BRBS channel. The divergent, linearly polarized output of an 830-nm laser diode is converted to collimated, circularly polarized light, converged through a clearance hole in a flat, 45-deg mirror, then further shaped to fill a concave, tilted, spinning mirror. The spinning mirror converts the stationary source to a circularly scanned point of light surrounding the hole in the 45-deg mirror, which is then directed toward the exit pupils by the 45-deg mirror. The scanned circle of light subtends an angle of approximately 3 deg at each eye of the subject. The projected images of the two patches of light falling on the spinning mirror create the exit pupils, located at the subject’s eyes, 40 cm from the center of the 45-deg mirror.

Figure 1

Ray tracing diagram and physical layout of the BRBS device: PD=photodiode; PBS=polarizing beamsplitter; QWP=quarter wave plate; and BS=beamsplitter.

Reflected light bundles from the retinas of both eyes follow the same paths back through the clearance hole until they are passed to detectors by a plate beamsplitter (the detector assembly has been rotated 90 deg about the optical axis for clarity of illustration). A knife-edge reflecting prism separates the spatially preserved signals from the right and left eyes. Polarizing beamsplitters separate the light bundles for each eye into orthogonal components of polarization (X and Y), which are separately converged onto two photodetectors conjugate to the retina to enable differential polarization detection, as described previously.⁴ The laser outputs of 0.20 mW/cm² at the subjects’ pupil are well below the established ANSI Z-136 safety standards of 0.56 mW/cm² for an indefinite period.

For data acquisition, subjects are seated at the location of the exit pupils, which are each 40 mm square and separated by 10 mm between the nasal edges. This geometry allows for the interpupillary separation of adults and children. Positioning of this hand-held instrument is facilitated by a triangulation, range-finding technique that utilizes two laser pointer modules projecting from above and below toward a point 1 cm anterior to the midpoint between the two exit pupils. The red laser spots were slightly separated horizontally to provide the needed direction clue for positioning forward and back relative to the subject.

2.2.

Optical Design: Focus Detection System

The concept of the monocular focus detection system has been described previously.^{5 7} The critical component of the apparatus is a bull’s-eye photodetector with two concentric active surfaces of equal area, a central circle and a surrounding annulus, positioned optically conjugate to a point light source. Using this configuration, sharply focused light falls only on the central area, while defocused light falls on both the central area and annulus, so that the ratio of center/annulus increases as the focus of the optical system improves.

The design of the binocular focus detection system is detailed in Fig. 2. A 785-nm laser diode produces monochromatic, diverging, linearly polarized light oriented vertically. A lens collimates the divergent light, which then passes into a polarizing beamsplitter. The beamsplitter is oriented to reflect essentially all of the vertically polarized light toward the subject. Light then passes through a quarter wave plate to produce circularly polarized light. A second lens converges the light to a point focus at the center of the hole in the flat 45-deg mirror (omitted from Fig. 2 for clarity; see Fig. 1). This places the 785-nm point source image 40 cm from the subject’s eyes. The small amount of light reflected by the fundus maintains most of its circular polarization but changes handedness on reflection before exiting the eye and returning to the quarter wave plate. The reflected light emerges from the quarter wave plate horizontally polarized and is passed toward the photodetectors by the polarizing beamsplitter.

Figure 2

Focus detection system (viewed from above): light emerges from the laser diode and travels toward the subject. The subject fixates on the image of the point source laser diode. Only the return light path is shown for clarity. Drawing not to scale.

A near-infrared wavelength was selected to minimize reflex pupillary constriction and thus loss of signal power.⁸ To separate the optical pathways of the focus detection system (785 nm) from the BRBS (830 nm), appropriate bandpass filters were inserted before the respective detectors, and the optics were coated with appropriate antireflection coatings. To electronically filter the desired focus detection reflected signal from background noise, a driver modulated the laser diode to produce a square wave at 400 Hz.

The circular polarization system was designed to detect maximum light reflected from the fundus while removing half of the depolarized light (produced by facial reflections and other diffuse reflections) from the signal. As long as a substantial fraction of the circularly polarized light reflected from the retina retains its polarization, this optical arrangement maximizes the light returning to the bull’s-eye detector.⁹

The detectors and laser diode were aligned using parallax to locate them in conjugate planes and were then visibly centered under high magnification. Exact optical conjugacy of the point source and detector was essential for proper performance, ensuring that as long as the eye was focused, the retina remained optically conjugate to both.

2.3.

Electronics and Signal Acquisition

The signals were acquired, processed, analyzed, and stored in near-real time in a custom virtual instrument environment developed using Lab Windows software (National Instruments, Austin, Texas). For each subject, a preliminary scan was obtained with eyes closed, and this background scan was then digitally subtracted from all subsequent data scans, improving the signal-to-noise ratio.

The focus detection laser light flashed on and off at 7 Hz to attract the subject’s attention. The background-subtracted output of the bull’s-eye photodetectors was displayed as calculated power spectra of the center (C) and annulus (A) signals. To adjust for potential differences in overall reflectivity of the eye, the ratio C/A was calculated, producing a minimum value of C/A=1 for a defocused eye. This ratio allowed focus detection system readings to be compared among subjects with varying reflectivity. As the value of A approached zero, the ratio C/A approached infinity. Therefore, the normalized ratio (C−A)/(C+A) was also calculated and stored. This ratio produced a predictable range of the focus detection system (FDS) output from 0 for a defocused eye to 1 for an ideally focused eye.

BRBS data were collected only during the “off” portions of the focus detector duty cycle to minimize interference. The software computed the percentage of power at 100 and 200 Hz, and characterized a reading as central fixation (200 Hz>100 Hz) or paracentral fixation (200 Hz<100 Hz). A reading was characterized as bilateral (both eyes with central fixation), unilateral (one eye with central fixation), or no reading (neither eye with central fixation).

2.4.

Device Operation

The prototype portable BRBS is shown in Fig. 3. The scanning laser beams exited the instrument toward the eyes through the exit hole in the face of the device [Fig. 3(b)]. Cables (not shown) connected power and data lines from the scanner to a “lunch box” data acquisition computer. The subject was seated across from the examiner, without head restraint. Room lights were dimmed to enhance interest in the fixation target. A blinking fixation target, the focus detection laser diode, was presented in combination with a synchronized beeping tone to draw the subject’s interest to the instrument [Fig. 3(b)]. A single reading was obtained in <0.5 sec. The software was configured to obtain a rapid sequence of 0.5-sec readings over 5 to 10 sec.

Figure 3

Prototype Pediatric Vision Screening device. (a) Illustration of device from operator’s side. (b) Subject’s view of fixation target (focus detection system) surrounded by faint ring (the scanned BRBS laser). The intensity of the ring is artificially enhanced in this photograph.

In anticipation that the PVS will be used as a screening tool, the software illuminated an array of four light emitting diodes (LEDs) mounted on the instrument [Fig. 3(a)]. The LEDs indicated right-eye alignment, right-eye focus, left-eye focus, and left-eye alignment, respectively. Each LED indicated green for a pass measurement, red for a refer measurement, or dim for an inconclusive reading, with thresholds set in software as described later. A single red indicator at the end of a series of measurements would, in practice, indicate the need for a comprehensive eye examination, while any dim indicator would indicate a need to repeat the screening. Four green LEDs would indicate the desired state, namely bilateral, simultaneous foveal fixation with bilateral simultaneous focus.

The front panel of the custom-programmed Lab Windows panel is shown in Fig. 4. The software displayed the results of the data analysis in both graphical form on the computer screen and also by illuminating the array of four LEDs described earlier. The fast Fourier transform (FFT) power spectrum of the BRBS signals was displayed as separate plots for the right and left eyes, and the FFT power spectrum of the focus detection signals was displayed as a pair of peaks for each eye—one representing the center (C) of the bull’s-eye photodetector, the other the annulus. Red and green lights in the upper right-hand corner of the RBS and focus detection (FD) plots indicated the status of the red and green LEDs on the instrument.

Figure 4

Pediatric Vision Screener software, main panel. Left side shows acquisition and control parameters. Panels show RBS (upper) and focus detection (lower) output. Upper light within each panel indicates passing measurement. In this case the subject demonstrated bilateral foveal fixation and bilateral focus.

2.5.

Volunteer Testing

To determine whether the PVS could identify strabismus and/or poor focus, measurements were obtained from one control and one strabismic volunteer. An orthoptist performed a “gold standard” examination, measuring best-corrected visual acuity, refractive error, binocular vision, and ocular motility. Residual refractive error was measured with vision correction in place if worn. The study was approved by the appropriate institutional review boards, and informed consent was obtained from all subjects.

The control subject was female, age 32, with hazel eye color and uncorrected visual acuity measured as 20/15 right eye (RE) and 20/15 left eye (LE). The strabismic subject was also female, age 35, with brown eyes and refraction RE: −3.25+0.75×180, LE: −4.50+1.50×002. With contact lenses there was no residual refractive error, and corrected visual acuity measured 20/60 in the amblyopic right eye and 20/25 in the left eye.

Subjects were asked to fixate centrally with both eyes, then centrally with each eye separately as the untested eye was covered with a clinical occluder. Subjects were then asked to fixate centrally and in four ordinal directions on the ring of laser light alternately (1.5 deg from the center). This protocol allowed testing for repeatability, detection of cross talk between channels, and identification of false positive responses. Five readings were obtained for each direction of fixation.