VIRTUAL RETINAL DISPLAYViirre, E., Pryor, H., Nagata, S., Furness, T. A. (1998). The Virtual Retinal Display: A NewTechnology for Virtual Reality and Augmented Vision in Medicine. In Proceedings of MedicineMeets Virtual Reality, San Diego, California, USA, (pp. 252-257), Amsterdam: IOS Press andOhmsha.

The Virtual Retinal Display: A NewTechnology for Virtual Reality andAugmented Vision in Medicine.Erik Viirre M.D. Ph.D. Homer Pryor, Satoru Nagata M.D. Ph.D.and Thomas A. Furness III Ph.D.Human Interface Technology Laboratory, University of Washington.Box 352142 Seattle WA 98195-2142AbstractIntroduction: The Virtual Retinal Display (VRD) is a new technologyfor creating visual images. It was developed at the Human InterfaceTechnology Laboratory (HIT Lab) by Dr. Thomas A. Furness III.The VRD creates images by scanning low power laser light directlyonto the retina. This special method results in images that are bright,high contrast and high resolution. In this paper, we describe howthe VRD functions, the special consequences of its mechanism ofaction and potential medical applications of the VRD, includingsurgical displays and displays for people with low vision. Adescription of its safety analysis will also be included. In one set oftests we had a number of patients with partial loss of vision viewimages with the VRD. There were two groups of subjects: patientswith macular degeneration, a degenerative disease of the retina andpatients with keratoconus. Typical VRD images are on the order of300 nanowatts. VRD images are also readily viewed superimposedon ambient room light. In our low vision test subjects, 5 out of 8subjects with macular degeneration felt the VRD images were betterand brighter than the CRT or paper images and they were able toreach the same or better level of resolution. All patients withKeratoconus were able to resolve lines of test several lines smallerwith the VRD than with their own correction. Further, they all feltthat the VRD images were sharper and easier to view. The VRD is asafe new display technology. The power levels recorded from thesystem are several orders below the power levels prescribed by the

American National Standard. The VRD readily creates images thatcan be easily seen in ambient roomlight and it can create images thatcan be seen in ambient daylight. The combination of high brightnessand contrast and high resolution make the VRD an ideal candidatefor use in a surgical display. Further, tests show strong potential forthe VRD to be a display technology for patients with low vision.1. IntroductionThe Virtual Retinal Display (VRD) is a new technology for creating visual images. It wasdeveloped at the Human Interface Technology Laboratory (HIT Lab) by Dr. Thomas A. FurnessIII. The VRD creates images by scanning low power laser light directly onto the retina. Thisspecial method results in images that are bright, high contrast and high resolution. Currentprototypes of the system produce full color images at a true 640 by 480 resolution.The technologies of virtual reality (VR) and augmented reality (AR) are the new paradigmfor visual interaction with graphical environments. The features of VR are interactivity andimmersion. To achieve these features, a visual display that is high resolution and wide field ofview is necessary. For AR a visual display that allows ready viewing of the real world, withsuperimposition of the computer graphics is necessary. Current display technologies requirecompromises that prevent full implementation of VR and AR. A new display technology called theVirtual Retinal Display (VRD) has been created. The VRD has features that can be optimized forthe human computer interfaces.The VRD is a visual display device that uses scanned light beams. Instead of viewing ascreen, the user has the image scanned directly into the eye. A very small spot is focused onto theretina and is swept over it in a raster pattern. The VRD uses very low power and yet can be verybright. The technology has been developed such that the scanning element will cost only a fewdollars in mass production. Low cost light sources, optics and controllers will make up the rest ofthe system. Ultimately, the overall device should be very inexpensive yet it will be small enough tomount on a spectacle frame.The development of this device has been driven by the need for a ubiquitious display that islightweight, full color and high resolution. In particular, the demands for displays for virtualenvironments and augmented vision are most pressing. In the past, virtual environments displayshave been very heavy, low resolution and have a small field of view. To create compelling virtualenvironments, the opposite is needed. The demands of displays for augmented reality, where thecomputer graphics image is superimposed on the real world, include a bright, high contrast image,and color that is appropriate. For example, an augmented vision display for a surgeon, whichmight provide him anatomic navigation information, would need to be unobtrusive during most ofthe procedure, produce bright enough images to be seen under the lights of the operating theatreand have color matched images that correspond to what the surgeon is seeing. The specialcharacteristics of images from the VRD may make it very useful for people with partial loss ofvision.Figure 1 is a block diagram of the VRD. Laser sources are introduced into a fiber optic

strand which brings light to the Mechanical Resonance Scanner (MRS) (patent pending). The MRSis the heart of the system. It is a lightweight device approximately 2 cm X 1 cm X 1cm in size andconsists of a polished mirror on a mount. The mirror oscillates in response to pulsed magneticfields produced by coils on the system mounting. It oscillates at 15 KHz and rotates through anangle of 12 degrees. The high frequency of scanning allows the fine resolution in the imagesproduced. As the MRS mirror moves, the light is scanned in the horizontal direction. Because themirror of the MRS oscillates sinusoidally, the scanning in the horizontal direction has beenarranged for both the forward and reverse direction of the oscillation. The scanned light is thenpassed to a mirror galvanometer or second MRS which then scans the light in the verticaldirection.The horizontally and vertically scanned light is then introduced to the eye. The light canbe sent through a mirror/combiner to allow the user to view the scanned image superimposed onthe real world.LASERHorizontalScannerVerticalScannerDeliveryOpticsVGAInputControlling ElectronicsModulatorFigure 1. Block Diagram of VRD systemsVRD versus Pixel Based DisplaysThe mode of illumination of the retina by the VRD is quite different from conventionalscreens. The scanning mechanism rapidly sweeps a spot of light over the retina. The spot passesover the retinel (an area analogous to the retinal area where a pixel is focused ). Thus the retinel isnot illuminated uniformly in time. Further, the actual time of illumination is extremely brief (40nanoseconds). There is only a brief spike of illumination of a portion of the retina for each refreshcycle of the display. The light from the VRD is coherent and very narrow band in wavelength. TheVRD can be configured such that the spot actually overlaps retinels or is smaller than a retinal area.Table 1 summarizes the differences between the pixel based display and the VRD.Table IPixel Based Display VRDIllumination constant over whole pixel Light scanned across retinapersistent light emission Short transient light emissionnon coherent light coherent lightbroadband color narrow band color

pixels separated by mask Spot can overlap retinels in scansAids for the Partially Sighted: The NeedPeople with partial vision constitute approximately 2% of the population in King Countyaccording to an internal study by Seattle Community Services for the Blind andPartially Sighted. Other centers around the country find similar population numbers[1-3]. Thepartially sighted have several major needs. According to CSBPS, sufferers with low vision mostoften request aids to allow them to read text or watch television and they may also require aids fornavigation or other activities of daily living. Current visual aids include simple glass magnifiers,video magnifiers and custom computer display enhancers. However current devices use olddisplay technologies. The old displays have inherent limitations in resolution, brightness, contrast,and field of view[4-8]. Further, the older display technologies are generally not portable and haveinherent characteristics that make them clumsy for use by the partially sighted.In this paper, we will describe potential medical applications of the VRD, includingdisplays for people with low vision and surgical displays. A description of its safety analysis willalso be included.2. MethodsFor our safety analysis, we measured the light power output of the VRD when it wascreating images. We had subjects adjust the brightness of the VRD images in a see throughconfiguration that allowed them to see an image on a conventional CRT screen. The VRD imagebrightness was adjusted so that it appeared equal to the brightness of the CRT images.The tables below show the results of some trial tests of low vision subjects with the VRD.In these tests subjects were brought in and gave informed consent. They were shown a series ofvision test images on paper, a computer screen and with the VRD. Their visual acuity was testedwith a standard office vision chart. For each display they were then shown test images to determinetheir resolving ability (acuity) and if any distortions were present (astigmatism or linear distortionson an Amsler grid) The performance on each medium was recorded and the subject’s subjectiveimpression of the visual image was also determined. The prototype VRD system was used forthese tests.In our pilot study we did a straightforward comparison of image quality of images from theVRD and a CRT and a images on paper. We controlled angular size of the images to be able tocompare best visual acuity. Image intensity was not controlled.Acuity measures: Landolt C’s.Image distortion Measures: Astigmatism stars and Amsler grids.Subjective impressions of the images.Subjects: Normal, Macular Degeneration, Keratoconus3. ResultsIn our safety analysis, all subjects were readily able to match the VRD brightness to thebightness of the control images. Power output values of the VRD varied from 50 to 1200nanowatts. Typical VRD images are on the order of 300 nanowatts. Typical VRD images are alsoreadily viewed superimposed on ambient room light. Normal subjects are all able to see VRD

images clearly. All 8 formally tested subjects were able to resolve VRD targets within one line ofCRT or paper targets. 4 were able to resolve targets at the same resolution. 5 of 8 normal subjectsreported VRD images to be “as sharp” or “sharper” than CRT or paper targets. There was nodistortion detected with astigmatism stars or Amsler grids.Macular Degeneration SubjectsMD subjects generally saw VRD targets as well subjectively and objectively as the CRT andpaper targets. Macular degeneration is a degradation of the visual receptors in the central part of theretina resulting in a decreased ability to read or recognize objects such as faces.Their visual acuitywas sharper with the VRD in some cases due to the pinhole effect on refractive error. Localizationof the small pinhole was difficult for some subjects.Keratoconus SubjectsKeratoconus is a distortion of the cornea. It results in blurred, defocussed images. Allkeratoconus subjects reported that they saw VRD images more sharply than any other visual targetsand in any viewing condition: no correction, glasses correction or contact lens correction (whichnormally provides the best vision). All subjects had equal or higher visual acuity with the VRDtargets, again even when wearing a gas permeable contact lens.Subject Visual Acuity Vision Disorder Paper CRT VRDOS OD OD OS OD OS OD OS1 20/400 20/200 Macular Degeneration 1/2 4 2/3 3 2/3 42 20/200 20/400 Macular Degeneration 3/4 1 3/4 1 5/6 1/23 20/800 20/800 Optic Neuritis 1 3 1 3 1 24 20/200 20/80 Macular Degeneration 1 6 1 6 0 65 20/400 20/400 Keratoconus 1 1 1 1 6 76 20/400 20/400 Keratoconus 1 1 1 1 6 77 20/400 20/400 Keratoconus 1 1 1 1 7 78 20/20 20/20 Retinal Pigmentosa 6 6 6 6 6/7 6/7Note: LegendVision test accomplished using Landolt C charts. 1 20/400Snellen Chart used to determine visual acuity. 2 20/200Amsler grid and Astigmatism tests showed no difference betweenthree conditions.3 20/1604 20/1205 20/806 20/407 20/30Augmented Reality.One of the leading applications for the VRD will be augmented vision andaugmented reality. Because of the bright images that can be produced by the VRD it will bepossible to use it in conditions as bright as ambient daylight. No current displays technology canproduce a portable image this bright. In augmented reality applications, images from the display are

overlaid on the real world for task enhancement. In augmented vision the images move with thesubject's head. In augmented reality, the images are held in registration with the real world as thesubject moves. For example, in an augmented reality application, a worker could see an instructionmanual or diagram overlaid on a part that is being repaired. Another use would be for peopleworking in environments with poor lighting conditions. The real world image could be enhancedelectronically and presented for better viewing with the VRD. In the elderly, opacities in the opticalmedia of the eyes increase glare as they view objects in sunlight or lighting conditions for night.The VRD could be used to image the world and then display it without the glare.Understanding of how the perception of images from the VRD interact with images fromthe real world is crucial for these applications. The test set ups for the beam characterization studiesand color perception studies will be ideal for augmented reality tests. In these tests the VRD willproduce images that will be superimposed on real world textures and backgrounds in a series oflighting conditions. The same image quality tests acuity, contrast, color and saturationdiscrimination will be performed while viewing the images with various backgrounds. The beamintensity will be varied by the subject to maximize viewing quality. Beam characteristics and colorsources will be reconfigured to maximize color contrast and hue matching with real world objects.4. ConclusionsThe VRD is a safe new display technology. The power levels recorded from the system are severalorders below the power levels prescribed by the American National Standard. The VRD readilycreates images that can be easily seen in ambient roomlight and it can create images that can be seenin ambient daylight. The combination of high brightness and contrast and high resolution make theVRD an ideal candidate for use in a surgical display. Further, tests show strong potential for theVRD to be a display technology for patients with low vision.Our future projects are:1.) Study the basic psychophysical processes of image perception from scanned lasersincluding resolution, contrast and color perception2.) Study the interaction of VRD images with images from the real world to enhance theaugmented reality applications of the technology.3.) study VRD image perception in partially sighted users.4.) design VRD light scanning paradigms to optimize image resolution, contrast in lowvision subjects.5.) Design text, image and computer icon representations for low vision users and test speed andaccuracy of recognition of those representations in the Seattle low vision population.5. AcknowledgementsThe development work for the VRD was funded by Microvision Inc. of Seattle. WA. VirtualRetinal Display and VRD are registered trademarks of Microvision. Perceptual research funding isbeing provided by the National Science Foundation Grant # IRI-9703598.

6. References[1] R. Robinson, J. Deutsch, H. S. Jones, S. Youngson Reilly, D. M. Hamlin, L. Dhurjon,and A. R. Fielder, “Unrecognised and unregistered visual impairment,” Br J Ophthalmol, vol. 78,pp. 736-40, 1994.[2] M. Yap and J. Weatherill, “Causes of blindness and partial sight in the BradfordMetropolitan District from 1980 to 1985,” Ophthalmic Physiol Opt, vol. 9, pp. 289-92, 1989.[3] J. M. Gibson, J. R. Lavery, and A. R. Rosenthal, “Blindness and partial sight in anelderly population,” Br J Ophthalmol, vol. 70, pp. 700-5, 1986.[4] J. Brabyn, “Problems to be overcome in high-tech devices for the visually impaired,”Optom Vis Sci, vol. 69, pp. 42-45, 1992.[5] B. Collins and J. Silver, “Recent experiences in the management of visual impairment inalbinism,” Ophthalmic Paediatr Genet, vol. 11, pp. 225-8, 1990.[6] C. C. Krischer, M. Stein Arsic, R. Meissen, and J. Zihl, “Visual performance and readingcapacity of partially sightedpersons in a rehabilitation center,” Am J Optom Physiol Opt, vol. 62,pp. 52-8, 1985.[7] A. G. Mathur, I. N. Raizada, R. Maini, and A. K. Maini, “Partially sighted--theirmanagement with low vision aids,” Indian J Ophthalmol, vol. 34, pp. 350-2, 1986.[8] O. Overbury, W. B. Jackson, and C. Hagenson, “Factors affecting the successful use oflow-vision aids,” Can J Ophthalmol, vol. 22, pp. 205-7, 1987.

Abstract—This pilot study examined the performance of analternative computer visual interface, the Virtual RetinalDisplay (VRD), for low-vision use. The VRD scans laser lightdirectly onto the retina, creating a virtual image. Since visuallyimpaired individuals can have difficulty using computer displays,a matched comparison study was done between the VRDand the standard cathode ray tube (CRT) monitor. Readingspeed and acuity tests were collected from 13 low-vision volunteersselected to represent the broad range of partially sightedindividuals actively involved in the work force. Forty-sixpercent of subjects had highest visual acuity while viewing theVRD; 30% of subjects had highest acuity viewing the CRT; and24% of subjects had equal acuity across the two displays.Although mean reading speed across all 13 subjects indicatedno significant difference between displays, individual subjectswith predominantly optical causes of low vision exhibited clin-431Journal of Rehabilitation Research and DevelopmentVol. 38 No. 4, July/August 2001Pages 431–442

The virtual retinal display as a low-vision computer interface:A pilot studyConor P. Kleweno, BS; Eric J. Seibel, PhD; Erik S. Viirre, MD, PhD; John P. Kelly, PhD;Thomas A. Furness III, PhDHuman Interface Technology Lab, University of Washington, Seattle, WA 98195-2142; Children’s Hospital RegionalMedical Center, University of Washington, Seattle, WA 98195-2142This material is based upon work supported by the National ScienceFoundation (award #9801294) and Howard Hughes Medical Institute.This study was presented, in part, on February 20, 1999 at the OpticalSociety of America (OSA) “Vision Science and Its Applications’’ TopicalMeeting, Santa Fe, NM.Address all correspondence and requests for reprints to: Eric J. Seibel, PhD,Research Assistant Professor, Mechanical Engineering, and Assistant Directorfor Technology Development at the Human Interface Technology Lab, 215Fluke Hall, University of Washington, Box 352142, Seattle, WA 98195-2142;email: eseibel@hitl.washington.edu.ically important increases in reading speed versus the CRT.However, most subjects with predominantly retinal damageshowed a slight disadvantage using the VRD. We give theoreticalexplanation to the bifurcated results and conclude that fora subset of low-vision users, the VRD technology is verypromising as a basis for future low-vision aids.Key words: computer displays, low vision, reading, retinal

scanning, visual disabilities, visual impairment, VRD.INTRODUCTIONIn recent years, the use of personal computers in theworkplace has become prevalent. This ubiquity is apparentin all age groups, even the older-aged ones. In fact,50.7 percent of American workers aged 50–59 years oldand 32.6 percent of American workers 60 years old andover use a computer at work (1). A small but significantportion of computer users in the workforce are individualswith low vision, many that are in these older agegroups. It is estimated that over 14,000 low-vision individualsare actively working in Washington State alone(J. Olson, Washington State Department of Services for432Journal of Rehabilitation Research and Development Vol. 38 No. 4 2001the Blind, personal communication). Individuals withlow vision required to work with computers can find thecurrent standard interface, a cathode ray tube (CRT) monitor,a hindrance to their productivity.Adaptations made by low-vision individuals to thestandard computer interface may have limited effectivenessand can even reduce job performance. For example,page navigation with the use of magnifiers canmake screen navigation tedious, actually slowing readingspeed and requiring a very large screen size (2). Inaddition, the common practice of increasing image sizeon the retina by moving the eyes closer to the displayscreen is a poor long-term solution, because of theinduced visual and musculo-skeletal strain (3). An alternativeapproach to simple enlargement is a fundamentallydifferent strategy that can provide increased visualacuity, hence requiring less magnification to read displayedtext. A possible alternative visual interface is theVirtual Retinal Display (VRD), which is the first retinallight-scanning system specifically intended for use as adisplay. The purpose of this study is twofold: (1) toevaluate the VRD as an alternative low-vision computerinterface and (2) to provide the theoretical motivationfor implementing the VRD (a retinal light scanningdevice) in the low-vision population. We compare theVRD with the CRT by utilizing visual acuity and readingspeed tests.The Virtual Retinal DisplayThe VRD scans modulated, low-power laser light toform bright, high-contrast, and high-resolution imagesdirectly onto the retina. This technology is related to thetechnology underlying the scanning laser ophthalmoscope(SLO) (4) but with the sole intent of image display,not acquisition. The VRD accepts the standard (RGBcolor or monochrome) output of a computer and generatesa raster-scanned image similar to the CRT monitor

(Figure 1). The portable VRD used in this study convertsthe VGA(red only) video output of a computer into a signalthat modulates the laser diode light source. The modulatedbeam of light (636 nm) is scanned horizontally(15.75 kHz) and vertically (60 Hz) by two mirrors. A lenssystem converges the raster-scanned beam to a 0.8-mmexit pupil. When the viewer aligns his or her eye at theexit pupil, the collimated beams of scanning light createa virtual image that appears in the distance. A detailedexplanation of the VRD during early development at theUniversity of Washington is found in Johnston andWilley (5).Retinal Scanning Technology in Low VisionScanned laser light for use in low-vision researchand rehabilitation has previously been consideredadvantageous by several authors (6–9). Laser sources,as used within the VRD, can produce images beyond thebrightness and contrast of conventional displays, suchas the CRT and liquid crystal display (LCD). For example,miniature LCD displays used for the purpose of awearable low-vision aid, project inferior images interms of contrast and brightness compared to the CRT(10). Since a scanned laser beam is capable of intensitybeyond what is safe for the human eye, the VRD hasbeen designed and shown (11) to produce images at safelevels, well below maximum permissible exposure levelsas defined by ANSI- and FDA-regulated standards.The capacity of a display to produce bright images isimportant for low-vision use. Cornelissen et al. (12)tested partially sighted individuals with a wide range ofmaladies and found significant visual acuity improvementat higher illuminance levels. Specifically, higherilluminations have been suggested to improve readingspeed for patients with macular degeneration (MD)(13).Previous research with low-vision individuals viewingretinal scanned images has shown promising resultsin the clinic. Webb and Hughes (6) reported dramaticimprovement in visual acuity (up to 20/70) for severalpatients who previously could only distinguish light fromdark. Culham et al. (14) used a SLO in low-vision readingperformance testing to both locate and display virtualimages onto optimal retinal locations for reading. Theseauthors suggest that their methods could be used to teachlow-vision patients (e.g., with MD) how to more effectivelyuse the remaining functional areas of their retina.Figure 1.Schematic diagram of the portable VRD used in this study. The directlymodulated laser diode produced a monochrome red (636 nm) output.The video input was standard VGA (6403480) color video froma PC laptop computer.433

KLEWENO et al. VRD as a low-vision computer interfaceIn addition, visual acuity and survey data from eight lowvisionsubjects comparing VRD and CRT images with theuse of full-color display systems have been reported byViirre, et al. (15). In all of these studies, the unique capabilitiesof retinal light scanning benefited individualswith low vision.However, these studies used large, sophisticatedlab systems, impractical for use as low-vision aids.Also, in the Viirre et al. (15) investigation, displaybrightness was optimized for each individual and wasnot matched in a controlled comparison. In our study, aportable, monochrome red version of the VRD is used tobetter simulate a low-vision aid. We also offer the firststudy to match luminance and field of view (FOV)between the VRD and the CRT. Our research goal wasto conduct a controlled, quantitative performance comparisonbetween the portable VRD and standard CRT interms of visual acuity and reading speed for individualsin the work force over a wide range of low-visionconditions.METHODSSubjectsOur study included 13 individuals with low vision,and was conducted at the Washington State Departmentof Services for the Blind (WSDSB). All subjects gaveapproved, informed consent. The 13 low-vision volunteerswere recruited by the WSDSB to comprise a varietyof vision conditions (Table 1). All 13 individuals, exceptone, were either actively employed or in graduate school.The subjects ranged in age from 28 to 59 years old witha mean age of 41.2 years old (SD510.3). Each subjectwas surveyed regarding his or her vision history and currenteye condition, and each survey was verified frompatient records at WSDSB.Materials and ApparatusWe measured visual acuity using Landolt “C’s,” followingthe specification that the width of the “C” and thegap in the “C” are one-fifth the dimension of the height.The “C’s” on the acuity chart had a range of sizes of 4.7ºto 0.3º, corresponding to a 20/1130 to 20/70 visual acuityrange. Acuity was measured while subjects viewed theCRT with white on black color contrast and the VRDwith red on black color contrast. The VRD was set at itshigher light output power level, matching the CRT interms of retinal illuminance (see below).A unique reading speed test based on the MinnesotaLow-Vision Reading test (MNRead™) (16) was used tocompare performance between the CRT and the VRD.Rather than implementing scrolled text or Rapid SerialVisual Presentation (RSVP) to overcome FOV constraints,

we chose to design entire words that more closely simulatedthe selective reading involved in actual computing. Thewords were presented in an unrelated manner becauseLegge et al. (16) have shown that reading speed of unrelatedwords correlates directly with normal sentence reading.In testing, three words at a time were presented to the subject:one five-letter, one four-letter, and one three-letterword. The three words were on three separate lines withinthe display field. The order (3-, 4-, or 5-letter) and placement(top, middle, bottom) were randomized. For each subject,no group of three words appeared together more thanonce. Individual words appearing more than once was rare(frequency 1 percent). Lists of words were randomizedwith the four test conditions (see Procedure).The portable VRD displayed the red component of astandard VGA (6403480) color video signal using adirectly modulated laser diode (636 nm). A PC laptopcomputer produced the VGA output at 6403480 resolution.Sans serif Arial type font was chosen, because itallowed more letters to fit on the portable display, whichhad a measured FOV of 33º horizontal by 26º vertical.The VRD can have a much wider FOV, but this particularversion was calibrated with those parameters.Subjects’ heads were kept stationary during testing byusing a chin rest constructed in the lab (Figure 2).A 17-in. CRT (EIZ0 Flexscan TX-C7, Nanao Corp.)was used for the comparison testing. Its full white-screenluminance was 107 cd/m2. All illuminance and luminancelevels were measured using a Spectrascan PR-650spectrophotometer (Photo Research, Inc.). In addition,the spectrophotometer was used to measure the peakwavelength of the CRT red screen (628 nm) and the VRD(636 nm). The contrast and power levels for laser powerwere measured using a silicon diode optical meter (model1835-C, Newport Corp.). The Michelson contrast,(Lmax2Lmin)/(Lmax1Lmin), of both the VRD and CRT letterswas 0.99. Two power levels of the VRD were usedduring the experiment, averaging 1.27 mW (SD50.05),and 2.45 mW (SD50.14). These two power levels weremeasured when the VRD was illuminating its entire FOVand was at maximum contrast. The higher power level(2.45 mW) was set to match the luminance value of thefully illuminated CRT screen when predicting a pupil sizeof 3.3 mm (2.43 mW, see below). The lower power level434Journal of Rehabilitation Research and Development Vol. 38 No. 4 2001(1.27 mW) was set at half the luminance level of the CRTscreen (1.22 mW, see below), approximating the lowerbrightness red on black CRT display condition.To match the VRD with the CRT, the luminance ofthe fully illuminated CRT (LCRT5107 cd/m2) was converted

to the lumens captured by the eye, then convertedto watts (WVRD) by the following relation (where 1 cd51lm * sr21) (17):In this equation, A is the area of the CRT screen(0.070 m2), r is the steradian measurement (4.6 *1025) for a viewing distance of 432 mm and pupildiameter of 3.3 mm, and ?s is the radiometric conversion,assuming a photopic curve (142 lm/W at 636nm) (17). A pupil diameter of 3.3 mm was estimatedusing the empirical equation of pupil size as a functionof luminance and FOV reported by Stanley andDavies (18). This estimate of pupil size was confirmedby taking measurements of each subject’s pupilusing a semi-circular pupil diameter gauge, whichresulted in finding an average pupil diameter of 3.2mm, (SD50.72). To calculate the lumens captured bythe eye viewing the CRT screen as an extendedsource, the assumption was made that the 3.3-mmdiameter pupil captures the same solid angle fromeach pixel point source. Retinal illuminance wasequivalent between the CRT and the VRD because wematched fields of view for each display during the calibrationsand subsequent testing.Figure 2.Picture showing author demonstrating experimental setup, viewingthe VRD with his left eye. Subjects held their head in the chin rest asshown and viewed either the VRD or the CRT screen. The CRT screenshown in author’s line of sight was covered during actual testing. Thelaptop shown in the foreground was the device subjects used to manuallyadvance through the reading speed test.Table 1.Subject characteristics.Subject Age Diagnosis (of eye tested) ClassificationA5 58 Amblyopia, nystagmus Ambloyopia/restrictedA6 39 Congenital cataracts, surgical aphakia, congenital Optical/restrictedglaucoma, retinal scarsA7 54 Severed blood vessel, glaucoma Retinal/fullA8 59 Amblyopia, strabismus Amblyopia/fullA9 30 Surface wrinkling retinopathy, retinal scars, astigmatism Retinal/fullA10 28 Unknown FullA11 40 Retinal detachment, congenital cataracts, artificial IOL Retinal/restrictedimplantA12 36 Cataract, artificial IOL implant, floaters Optical/restrictedA13 32 Congenital aniridia, cataracts, corneal ulcers Optical/fullA15 45 Diabetic retinopathy, surgical aphakia Retinal/restrictedA16 33 Diabetic retinopathy, retinal detachment and scarring Retinal/restrictedA17 38 Glaucoma, congenital cataracts, surgical aphakia Retinal/restrictedA18 43 Glaucoma, congenital cataracts, nystagmus Retinal/fullNote: Both primary and any secondary (listed in such order) diagnoses are outlined for each subject. Subjects were classified by their primary eye condition intothree categories: optical, retinal, or amblyopic. The subjects were further classified as having “full” or “restricted” field of view. Restricted field of view describesany occlusion or visual field degeneration.WVRD(watts)5[L(cd/m2

CRT*A(m2)*r(sr)]/s(lm/watt)52.43mW[1]435

KLEWENO et al. VRD as a low-vision computer interfacePROCEDURETesting was conducted in a naturally lighted room.Ambient illumination in the testing room was controlledby measuring the luminance from a covered CRT screenbefore each subject test (53.10 cd/m2, SD53.93). Mylarshades were adjusted at these times to keep conditionsconstant for all subjects. The vision testing was donemonocularly; each subject was asked to choose his or herpreferred eye. The same eye was used for both the acuitytests and the reading speed tests. Throughout testing, subjectswore their own optical correction (contact lenses oreyeglasses) according to what they would normally wearwhen viewing a CRT screen at a distance of 17 in.Visual acuity testing was always completed beforethe reading speed tests. The order of acuity testing, interms of viewing the CRT or VRD first, was randomized.Each subject’s acuity was scored at the last line at whichhe or she correctly identified at least 3 of the 5 Landolt“C’s.”Subsequently, each subject was given a readingspeed test comprised of four different test conditions: (1)viewing a CRT with white letters on a black background,(2) viewing a CRT with red letters on a black background,and (3) viewing a VRD image with a lower power setting(1.27 mW), and (4) viewing a VRD image with a higherpower setting (2.45 mW). The CRT red-on-black contrastcondition was used to more closely match the CRT wavelengthwith the monochrome red VRD. The two powersettings of the VRD were used to determine if there wasan effect of retinal illuminance. The order of test conditionwas randomized.Each reading speed test (at each of the four test conditions)included four character sizes, 3.15º, 1.88º, 1.22º,and 0.74º, measured using a lower case “x” as reference.These four values were selected to comprise themidrange sizes between the intended subjects’ predictedlower reading threshold and their predicted upper plateauof reading speed. Reading speed, with respect to charactersize, has been empirically shown to start at a certainvalue, increase, then plateau, and then finally decrease, ascharacters become impracticably large, for both normaland low-vision subjects (19). We targeted the midrangebetween the subjects’ visual acuity threshold and plateau,predicting that reading rate would be most sensitive tothis range of character size. The character size range wasselected so as to encompass the predicted visual acuitylimits of our subjects. This prediction was based on subjects’records at the WSDSB.At the start of each reading test, subjects read at thelargest character size, then at successively decreased

sizes. For each character size, three (20-s) trials were conductedconsecutively. The subject manually advancedthrough the three-word presentations by clicking themouse button on the laptop and reading the words aloud.The number of correctly read words per 20-s trial wasrecorded; the subjects were allowed a brief rest inbetween each trial. In the few cases where the subjectreported fatigue, trial number was reduced from 3 to 2,since the subject’s first two scores were similar. At theconclusion of the reading tests, we asked subjects to ratethe VRD as “better, the same, or worse than the CRT’’ interms of perceived brightness and perceived clarity.RESULTSThe results for 13 of 15 low-vision volunteers arereported here. Subject A4 requested to be withdrawnfrom the testing because of fatigue. Subject A14 wasunable to locate the VRD exit pupil and maintain a stableimage on the small functional portions of peripheral retina.Therefore, we removed these two subjects from thestudy.Forty-six percent of subjects had highest visual acuitywhile viewing the VRD (at the increased power); 30percent of subjects had highest acuity viewing the CRT(white-on-black contrast); and 24 percent of subjects hadequal acuity across the two displays (Table 2). However,a repeated-measures t-test showed no significant differencebetween the two displays across all subjects(p50.24). A Pearson correlation test showed no significantcorrelation between improved acuity and improvedreading speed (r50.26; p>0.05).Individual reading performances are summarizedand listed in Table 2. Reading speed columns contain themean reading rates in words per minute (wpm) averagedover trial and the two respective VRD or CRT conditions.The fourth column over from the left shows both the subjects’minimum angle of resolution (MAR) and the onetesting character size that was chosen to average acrossfor each subject. This character size was one of the foursizes used in the testing (3.15º, 1.88º, 1.22º, and 0.74º)and was determined by multiplying their measured CRTMAR by five and then selecting the character size thatwas the next largest. For example, subject A5 had a MARof 0.24º (visual acuity520/290), therefore the 1.22º charactersize was chosen. The factor of five derives from436Journal of Rehabilitation Research and Development Vol. 38 No. 4 2001Snellen letter specifications, which require the smallestdetail in a Snellen letter be one-fifth the size of the overallletter. The Arial font we used displayed characterswith details that averaged 0.22 times the size of theoverall letter, thus approximating the Snellen specification.

In this way, reading performance could be evaluatedacross subjects relative to their own acuity limitations.Five subjects read faster while viewing the VRD,and eight subjects read faster viewing the CRT at theirrespective selected character size. Individual readingspeed averages ranged from 1813 (A13), to 242 percent(A15), when the VRD was compared to the CRT (Table 2).However, when averaging across all subjects, the twodisplays were almost equal in performance at 63.1 wpmfor the VRD versus 62.8 wpm for the CRT. A repeatedmeasurest-test showed that when all four font sizes wereincluded in the averaging, there was no significant differenceacross all subjects (p>0.5), which is shown graphicallyin Figure 3. Similarly, no significant difference wasfound (again using a repeated-measures t-test) in readingperformance between the two displays (p>0.5) whenusing only the selected character sizes.A summary histogram of the pilot data is shown inFigure 4. The average reading speeds at the selectedcharacter size for the 12 subjects categorized by theirlow-vision condition (subject A10 had an unknown eyecondition) is graphed. We classified these 12 subjects intothe following four categories as shown in Table 1: opticalcauses (A6, A12, A13), amblyopia (A5, A8), retinalcauses with full FOV (A7, A9, A18), and retinal causeswith restricted FOV (A11, A15–17). A restricted FOV isdefined here as a visual field with occlusion or visualfield degeneration. For the optical cause and the amblyopiagroups, the VRD (averaged over power levels) providedincreases in reading speed of 39.4 percent and 19.0percent, respectively, over the CRT (averaged over colorcontrast). Subjects with retinal causes plus restrictedfields of view, and subjects with retinal causes plus fullfields of view had greater reading speeds when viewingthe CRT than when reading the VRD (27.5 percent and17.1 percent, respectively). A two-way analysis of variance(ANOVA) of low-vision category and display conditionreading speeds showed no significant main effectfor category (F51.7, p>0.05), or for display condition(F50.24, p>0.05), and no significant interaction(F50.41, p>0.05). However, given the low sample size,analysis showed that the power was extremely low in thedifferent treatments (less than 0.3 at a50.05). Based onthe power analysis results, we would have needed about20 subjects per category in our study to achieve a significantpower level (e.g. 0.8) when comparing the lowvisioncategories.Table 2.Summary of individual performances.BetterSpeed Speed display

Acuity Acuity MAR/Char VRD CRT Percent (for Apparently ApparentlySubject VRD CRD (°) (wpm) (wpm) difference reading) clearer brighterA5 20/200 20/290 0.24/1.22 51.5 22.5 129 VRD VRD VRDA6 20/100 20/100 0.083/0.74 95.5 67.0 43 VRD VRD VRDA7 20/100 20/290 0.24/1.22 64.0 73.5 –13 CRT VRD VRDA8 20/70 20/70 0.058/0.74 111 114 –3 CRT VRD VRDA9 20/360 20/200 0.17/1.22 38.5 54.0 –29 CRT CRT VRDA10 20/140 20/200 0.17/1.22 57.0 56.0 2 VRD Same SameA11 20/140 20/200 0.17/1.22 18.0 11.5 57 VRD VRD VRDA12 20/100 20/70 0.058/0.74 94.3 113.5 –17 CRT VRD VRDA13 20/430 20/700 0.58/3.15 73.0 8.00 813 VRD VRD VRDA15 20/140 20/100 0.083/0.74 43.3 74.0 –42 CRT CRT VRDA16 20/200 20/290 0.24/1.22 68.0 101 –33 CRT Same SameA17 20/140 20/100 0.083/0.74 32.0 36.0 –11 CRT VRD CRTA18 20/100 20/100 0.083/0.74 74.5 86.0 –13 CRT VRD VRDAverage 63.1 62.8VRD and CRT reading speeds, listed in words per minute (wpm), are averaged over trial, character size and respective viewing condition (VRD or CRT). MARrefers to minimum angle of resolution for the CRT acuity test, and Char(°) refers to the next largest character size after five times the CRT MAR. Percent differencewas calculated as (SpeedVRD–SpeedCRT)/SpeedCRT. Apparently clearer and apparently brighter columns correspond to results of subject survey.437KLEWENO et al. VRD as a low-vision computer interfaceFigure 3.Average reading speeds for all 13 subjects. Reading speed scores are averaged across trial, type of display (VRD or CRT), and across subject.After averaging the reading speeds for the entire study, no significant difference is found between viewing VRD and viewing CRT.Figure 4.Reading speed summary for subject groups classified by eye condition (see Table 1 for classifications). Reading speeds are averaged over respectiveviewing condition (VRD or CRT). Subjects with optical causes of low vision showed greatest increases in reading speed (39.4%), and amblyopiasubjects had a 19.0% increase in reading speed when viewing the VRD compared to the CRT. For subjects with retinal causes of low vision,decreases in reading speed while viewing the VRD were found for both full FOV subjects (17.1%), and restricted FOV subjects (27.5%).Averaging across all subjects or within designatedcategories does not reveal the large differences in readingperformance between the displays that are apparent forindividual subjects. Seven subjects (A5, A6, A9, A11,A13, A15, and A16) demonstrated clinically importantreading speed differences between the two displays(define clinical importance as ±25-percent difference, E.Peli, personal communication). The reading speeds areplotted for four of these seven subjects (A13, A5, A9, andA16) at the four viewing conditions, each one representingone of the four low-vision categories (Figure 5).Subject A13 (optical causes, Figure 5A) had an 813-percentincrease in reading speed viewing the VRD at theselected character size of 3.15º. Note that this was theonly size where this subject was able to practically readCRT images and only in the white-on-black contrast.Subject A5 (amblyopia, Figure 5B) read 129 percentfaster with the VRD at the selected character size. In contrast,subject A9 (retinal full FOV, Figure 5C) had a 29percent decrease in reading speed when viewing the VRD

438Journal of Rehabilitation Research and Development Vol. 38 No. 4 2001compared to viewing the CRT, while subject A16 readCRT words 33 percent faster than VRD words (retinalrestricted FOV, Figure 5D). The remaining six subjects(out of the 13 total) showed comparable reading speedsbetween the VRD and the CRT throughout the four categories(25-percent difference).All subjects were surveyed with results listed inTable 2. Survey questions had the subjects compare thetwo displays in terms of the overall apparent brightnessand clarity of images. Nine of 13 subjects found VRDimages as clearer, while 2 of 13 found CRT images asclearer, and another 2 subjects rated both displays asequivalent in image clarity. In terms of brightness, 10subjects reported the VRD as having brighter images, 1subject found the CRT to be brighter, and another 2 subjectsrated both displays as equal in apparent brightness.A noteworthy result of the survey revealed that 12 of 13subjects expressed a difficulty or a specific dislike of redon black contrast. Statistical analysis demonstrated thatthe red on black color contrast had a significant detri-Figure 5.Average reading speed scores with respect to character size for subjects A13, A5, A9, and A16. In each graph, data points are connected by a bestfitsmoothed line. A: Reading speeds for subject A13. Individual had an 813% increase in reading speed when viewing VRD compared to viewingCRT at selected 3.15º character size. B: Reading speeds for subject A5. Subject had a 129% increase in reading speed when viewing VRDcompared to viewing CRT at selected 1.22º character size. C: Reading speeds for subject A9. Subject had a 29% decrease in reading speed whenviewing VRD compared to viewing CRT at selected 1.22º character size. D: Reading speeds for subject A16. Subject had a 33% decrease in readingspeed when viewing VRD compared to viewing CRT at selected 1.22º character size.A. B.C. D.439KLEWENO et al. VRD as a low-vision computer interfacemental effect on reading speed within the two CRT displayconditions when averaged across all subjects(p50.035).DISCUSSIONVisual AcuityAlthough visual acuity can be a poor predictor ofreading speed (20), it can be a measure of nonreadingperformance; for example, improved acuity may allowfor faster and more accurate recognition of graphical userinterface icons. The increase in acuity shown by 6 of 13subjects may be attributed to the VRD’s narrow exitpupil, which has two effects. The first is that the collimatedlight at the exit pupil produces a large depth offocus. The extended depth of focus allows retinal scanneddisplays like the SLO and VRD to remain unaffected bymoderate refractive errors in the eye, and in some lowvision

cases delivering an image of higher quality thanthe naturally apertured eye. The 0.8 mm diameter beamfrom the VRD is assumed to be diffraction limited at theretina. Thus, depth of focus is calculated for a perfectlyspherical wavefront (21) at 3 diopters, versus 0.2diopters for viewing the CRT. Interestingly, empiricalresults (not reported) have demonstrated that personsviewing the VRD do not notice refractive errors rangingup to a maximum of 6 to 12 diopters.The second effect of having the small exit pupil ishigher contrast because of less dispersive scattering by theoptical media (6). Damaged or abnormal media such asthe cornea, lens, or vitreous greatly add to the scattering ofincident light and hence decrease visual contrast as theindividual perceives the scatter as glare. Standard displayssuch as the CRT illuminate all optical media allowed bypupil size, including the abnormal media that cause scatter.The very narrow directed beam of scanned light allowsfor less diffuse scatter and possibly less interception ofscatter-causing media, producing a higher contrast retinalimage. Furthermore, a low-vision individual can strategicallyorient his/her eye with respect to the narrow laserbeam to minimize the glare from specific scattering points(e.g., corneal scars) to optimize image quality. These twoeffects may also explain why 69 percent of subjectsreported that the VRD displayed clearer images.Reading PerformanceFigure 3 shows the average reading speeds for allsubjects when viewing the VRD (averaged over bothpower levels) compared to viewing the CRT (averagedover both color contrast settings). This figure demonstratesthat mean reading speeds across all 13 subjectsindicated no significant difference between the displayswhen averaged over all font sizes. However, our subjectpopulation consisted of persons with varied vision conditionsand acuity limitations. Thus, selecting the mostappropriate character size for each subject as described inthe Results section allowed for a more reasonable comparisonacross subjects. Within each of the four lowvisioncategories, case-study analyses of subjects whoshowed clinically important differences in reading speedbetween displays (at the selected character size) provideinsight into the performance of the VRD versus the CRT.The selected character size for each subject appears tohave been in their own midrange of reading speed, basedon individual plots of reading speed versus character size.Subject A13, who was employed as a computer programmer,showed drastic improvement in acuity andreading speed while viewing the VRD compared to theCRT (Table 2, Figure 5A). This subject suffered fromlow vision due entirely to optical causes (aniridia,

cataracts, and corneal ulcers) and evidences the findingthat, in general, individuals with optically-based maladiesexhibited the greatest improvement in reading speedwhen using the VRD (39.4 percent, Figure 4). Analogousto improving acuity, we attribute this finding to theVRD’s narrow exit pupil and great depth of focus, as discussedearlier.Subject A5, who suffered primarily from amblyopia,showed higher acuity and an increase in reading speedviewing the VRD (Table 2, Figure 5B). The other subjectwith amblyopia (A8) demonstrated equal acuity and comparablereading speed between displays. Both subjects’results indicate that the VRD may offer a viable alternativefor individuals suffering from this condition (Figure 4).In contrast, subjects with retinal disorders showedhigher average reading performance viewing the CRT(Figure 4). For example, subject A9, who suffered fromsurface wrinkling retinopathy and retinal scarring, had a29 percent decrease in reading speed (Table 2, Figure5C) while viewing the VRD compared to the CRT.Although this subject’s macula was intact, she still indicatedthat the VRD was “difficult to see into’’ because ofbeing unaccustomed to restricting head movements.(Head-mounting the VRD may alleviate this problem forsimilar individuals, see below.)The VRD seemed to be even more problematic forsubjects with retinal causes and restricted fields of view.For example, subject A16 had a 33-percent decrease inreading speed at the selected 1.22º when viewing the VRD(Table 2, Figure 5D). This subject suffered from diabeticretinopathy, partial retinal detachment and retinal scarringbecause of retinal burns from previous laser treatments.The subject’s significant reading-speed decrease may beexplained by visual field defects since he reported that helost speed viewing the VRD because of losing the words inhis visual field. Thus, a limitation in the effectiveness ofscanning light technology may be found in individualswith retinal disorders that reduce their FOV.Most subjects (77 percent) perceived brighter imageswhen viewing the VRD. The VRD’s apparent brightnessmay be due, in part, to the Stiles-Crawford Effect (SCE)(22). The SCE describes the phenomenon that light sensitivityis optimal for light entering the eye near the center ofthe pupil and diminishes for light rays at the periphery ofthe pupil. Apparent brightness (luminous efficiency), basedon the SCE, can be quantified by considering a ratio inluminous efficiency of light with respect to that at the optimalentry point (optimal varies slightly per individual).Stiles and Crawford (22) presented the following empiricalformula to define h where p is a coefficient slightly dependenton wavelength (assume p50.05) and r is the radial

distance from the center of the pupil.Thus, the average luminous efficiency over agiven pupil diameter is calculated by integrating thefunction in two dimensions, producing a three-dimensionalluminous efficiency volume. For a calculated3.3-mm pupil viewing the CRT (see Methods), theluminous efficiency is calculated to be 86 percent. Areasonable pupil diameter of 5 mm (the largest pupilsize we measured) yields an average luminous efficiencyof 71 percent. These values are to be comparedto the smaller VRD exit pupil, where luminous efficiencyis 99 percent. The ratio of VRD efficiency tonatural pupil efficiency (when viewing a CRT display,for example) corresponds to a luminous efficiencyincrease of 15 percent (3.3 mm) to 40 percent (5 mm).The higher efficiency of the VRD’s 0.8-mm exit pupilbecause of the SCE is supported by our empirical comparisondata on perceived image brightness, as mostsubjects perceived the VRD as brighter than the CRTunder matched luminous conditions.There was no significant effect on reading speedbetween the two VRD power levels, which may be attrib-440Journal of Rehabilitation Research and Development Vol. 38 No. 4 2001utable to the wide variance of the optimum luminancelevel among individuals (12). The VRD can retain sufficientcontrast and resolution in images at lower brightnesslevels for those individuals highly sensitive to lightor glare. Some subjects (e.g., A13, Figure 5A) benefitedmore from the VRD lower power setting, a luminancehalf that of the white-on-black CRT.Most low-vision subjects (e.g., A9) reported difficultykeeping their eyes aligned with the VRD’s 0.8-mmdiameter exit pupil. The problem of losing alignment canoccur when head position moves relative to the displayand/or the eye scans within a wide field of view VRD.Recent studies with the portable VRD configured as ahead-mounted, augmented display has reduced exit pupilmisalignment during performance evaluations (23).Future research will attempt to alleviate the alignmentproblem in wider FOV retinal displays by adjusting themethod by which the beam enters the pupil. Thus, theprototype VRD using the disliked red on black color contrastand the experimental design of not configuring theVRD as head-mounted, produced an obvious performancebias against the VRD versus the CRT.Furthermore, recent research has confirmed that the redon black color contrast is not only least preferred but hasreduced reading rates versus green-, blue-, or white-onblackcolor contrasts (24).Another anticipated disadvantage of retinal light

scanning technology could be the appearance of flickerin scanned optical images. Retinal light scanning producesa pixel of essentially no persistence time versusthe well-known persistence of the CRT phosphor emission.However, research has indicated that human subjectsdo not detect flicker in the VRD any more than aCRT (25).Although we attempted to recruit subjects who wereactively employed and who encompassed a wide range ofvision maladies (Table 1), further research will be neededto address subjects with MD. An indication of theeffectiveness of retinal light scanning with MD subjectsis found in an earlier study by Viirre et al. (15). The fullcolorVRD demonstrated improved acuity in two of threeMD subjects versus the CRT, which was attributed to thebrighter, full-color VRD.In addition, retinal light scanning can be made bothsmall and inexpensive by the advent of new MEMS(microelectromechanical systems) optical scanners,miniature video cameras, and LED-based monochromeor multisource white light LEDs. A future display may beworn as a pair of oversized glasses or goggles to possiblyh5102pr2 [2]441KLEWENO et al. VRD as a low-vision computer interfaceeliminate pupil alignment problems and alleviate the poorergonomics of holding the head at a fixed position. Thesestudies are being carried out under the Universal Accessand Research to Aid Persons with Disabilities Programsof the National Science Foundation (NSF) to understandthe full potential of retinal light-scanning technologiesfor improving computer accessibility and navigationalability of all categories of low-vision users.CONCLUSIONWe compared the visual performance of low-visionsubjects when using a standard desktop CRT and a portable,monochrome red VRD. Most subjects showed equivalentor improved visual acuity when viewing the VRD comparedto the CRT. A reading performance test at controlledluminance levels matched to the CRT showed that on average,the VRD proved to be a comparable display. The greatestindividual improvements in reading speed with the VRDwere recorded with subjects having optical causes of lowvision. Retinal light-scanning displays such as the VRDoffer a safe visual interface to computers and are anticipatedto improve computer accessibility because of greaterapparent brightness, contrast, and depth of focus comparedto standard CRT monitors.ACKNOWLEDGMENTSThis work is in partial fulfillment of Conor Kleweno’shonors thesis in the Department of Bioengineering,

University of Washington. Conor Kleweno’s advisor isEric Seibel. Conor Kleweno is an accepted student atHarvard Medical School (2001), a recipient of the 1999Barry M. Goldwater National Scholarship, 2000Washington State Rhodes Scholarship finalist, and consecutiveHoward Hughes Research Fellowships (1998 and1999). We thank Bob Burstein for his technical assistance.We are indebted to John Olson and the staff at theWashington State Department of Services for the Blind,Seattle, WA. VRD and Virtual Retinal Display are trademarksof the University of Washington.REFERENCES1. US Census Bureau. Statistical abstract of the United States ofAmerica. Washington DC: US Census Bureau; 1999. p. 442.2. Beckmann PJ, Legge GE. Psychophysics of reading XIV. Thepage navigation problem in using magnifiers. Vision Res1996;36(22):3723–33.3. Jaschinski W, Heuer H, Kylian H. A procedure to determine theindividually comfortable position of visual displays relative to theeyes. Ergonomics 1999;42(4):535–49.4. Webb RH, Hughes GW, Pomerantzeff O. Flying spot ophthalmoscope.Applied Optics 1980;19:2991–7.5. Johnston RS, Willey SR. Development of a commercial retinalscanning display. Helmet- and head-mounted displays and symbologydesign requirements II. Proc of the SPIE 1995;2465:2–13.6. Webb RH, Hughes GW. Scanning laser ophthalmoscope. IEEETrans Biomed Eng 1981;28:488–92.7. Mainster MA, Timberlake GT, Webb RH, Hughes GW. Scanninglaser ophthalmoscopy. Ophthalmology 1982;89:852–7.8. Timberlake GT, Peli E, Essock EA, Augliere RA. Reading with amacular scotoma II. Invest Ophthalmol Vis Sci 1987;28:1268–74.9. Schuchard RA, Fletcher DC, Maino JM. A scanning laser ophthalmoscope(SLO) low-vision rehabilitation system. Clin Eye VisionCare 1994;6(3):101–7.10. Massof RW. Electro-optical head-mounted low vision enhancement.Pract Opt 1998;9(6):214–20.11. Viirre E, Johnston R, Pryor H, Nagata S. Laser safety analysis ofa retinal scanning display system. J Laser Appl 1997;9:253–60.12. Cornelissen FW, Kooijman AC, Dumbar G, Van der Wildt G,Nijland R. Illumination research as part of a visual assessmentof visually impaired individuals. Ophthalomologica1991;78:195–203.13. Eldred KB. Optimal illumination for reading in patients with agerelatedmaculopathy. Optom Vis Sci 1992;69(1):46–50.14. Culham LE, Fitzke RW, Timberlake GT, Marshall J. Use ofscrolled text in a scanning laser ophthalmoscope to assess readingperformance at different retinal locations. Ophthalmic Physiol Opt1992;12:281–6.15. Viirre E, Pryor H, Nagata S, Furness TA. The Virtual RetinalDisplay: a new technology for virtual reality and augmented visionin medicine. In: Proceedings of medicine meets virtual reality VI,studies in health technology and informatics. 1998;50:252–7.16. Legge GE, Ross JA, Luebker A. Psychophysics of reading VIII.The Minnesota low-vision reading test. Optom Vis Sci1989;66(12):843–53.17. Wyszecki G, Stiles WS. Color science: concepts and methods,quantitative data and formulae. (2nd ed.) John Wiley and Sons,Inc; 1982. p. 256–8, 265, 731.18. Stanley SA, Davies KA. The effect of field of view size on steadystate

pupil diameter. Ophthalmic Physiol Opt 1995;15(6):601–3.19. Legge GE, Rubin GS, Pelli DG, Schleske MM. Psychophysics ofreading II. Low vision. Vision Res 1985;25(2):253–66.20. Legge GE, Ross JA, Isenberg LM, LaMay JM. Psychophysics ofreading: clinical predictors of low-vision reading speed. InvestOphthalmol Vis Sci 1992;33(3):677–87.21. Hopkins RE, Stephenson D. Optical systems for laser scanners. In:Marshall GF, editor. Optical scanning. New York, NY: MarcelDekker; 1991. p. 38.22. Stiles WS, Crawford BH. The luminous efficiency of rays enteringthe eye pupil at different points. Proc R Soc Lond1933;B112:428–50.23. Lin S-K V, Seibel EJ, Furness TA. (accepted). Testing visualsearch performance using retinal light scanning as a future wearablelow vision aid. Int J Human-Computer Interact, two-volumeSpecial Issue on Mediated Reality.24. Seibel EJ, Gau CC, McQuaide S, Weghorst SJ, Kelly JP,Furness TA. Augmented retinal light scanning display for thelow vision: effect of text color and background on reading performance.Vision Science and Its Applications, OSA Technical442KLEWENO et al. VRD as a low-vision computer interfaceDigest (Optical Society of America, Washington DC, 2001);51–4.25. Kelly JP, Turner S, Pryor HL, Viirre ES, Seibel EJ, Furness III TA.Vision with a scanning laser display: comparison of flicker sensitivityto a CRT “Displays” (Submitted).Submitted for publication June 6, 2000. Accepted inrevised form November 11, 2000.