Go to previous article
Go to next article
Return to 2002 Table of Contents
Gregory L. Goodrich, Ph.D.
Psychology Service and Western Blind Rehabilitation Center
VA Palo Alto Health Care System
3801 Miranda Avenue
Palo Alto, CA 94304
Optical devices (OD) and video magnifiers (VM) have long been the standard reading devices prescribed for individuals with low vision. In the 1980's Optelec introduced the first head borne electronic aid  and in the past few years a number of companies have introduced black and white, as well as full color head borne units based upon CRT displays mounted in units similar to those used with some virtual computer games. These units can be used for reading and for (stationary) distance viewing. Their primary advantage appears to be their portability.
The most recent development in head borne devices has been the introduction of a virtual display using a monochromatic (red) laser [2, 3]. The Retinal Scanning Display(tm) (RSD) prototype used in this study was a monocular device placed in front of the dominant eye of the subject. The RSD uses a laser that safely projects images directly onto the retina of the eye. It provides a 27( X 20( field of view using a 650 nm (red) laser with a resolution of 800 X 600 pixels. The head borne unit weighs about 1 lb. The maximum brightness of the device is variable from 1 to 800 foot lamberts. The RSD was connected to a Telesensory Genie camera and X-Y table that provided the input to the RSD.
The primary goal of this pilot study was to determine how well low vision subjects read with the device in comparison to their best optical device and their prescribed VM. Of the 20 subjects studied, 18 were male and 2 female. Subjects ranged in age from 31 to 86 years of age with a mean of 70 years. Visual acuities ranged from 20/80 to 3/300 with a logMAR average of 1.125, although one subject reported a visual acuity of "counts fingers". Contrast sensitivity measured by the Pelli-Robson test averaged 0.98. One subject, possibly due to his particular form of color blindness, was unable to read with the RSD. All subjects but two were able to read visually with their existing devices. Of the two who could not, one was able to spot check mail with an illuminated magnifier, but not fluently read, and the other did not read at all. With the exception of these two subjects, five sessions of training was provided to each subject with each reading device (OD, VM, and RSD). This training helped ensure that all subjects were equally familiar with each device and eliminated potential practice effects in the comparison between the devices. Training sessions were 40 minutes in length and this limited the maximum reading durations reported here. Reading speed was measured using equivalent difficulty (5th grade) paragraphs of approximately 275 words. Duration was the total amount of time read during the final training session. Except where noted, the data is for the 17 subjects who read with all three devices.
The average reading speeds (upon completion of training) were: 49 words per minute (wpm) with the OD, 66 wpm with the VM, and 51 wpm with the RSD. Reading durations with the three devices were: 22 minutes with the OD, 38 minutes with the VM, and 31 minutes with the RSD. Thus, reading performance with the RSD, as measured by reading speed and duration, was midway between the conventional devices. Interestingly, the correlation between reading speed with the OD and RSD was not significant, while VM and RSD reading speeds were significantly correlated (r = 0.70), p < 0.05). For the two subjects unable to read with either conventional device, the RSD provided modest reading speeds of about 7 wpm and 26 wpm respectively. Both felt that their ability to read with the RSD was due to its clarity and brightness.
A series of forced-choice and open-ended questions assessed subjects' opinions of the VM and RSD. In general, most subjects preferred the VM, although all felt that the RSD provided the brightest and clearest image. Among the negative features of the RSD were its weight and the lack of comfort provided by the prototype's headband, which tended to slip down (and out of position) unless the mounting was very tight on the head. This lack of comfort contributed to the lack of acceptance of the RSD by most subjects. Also, most subjects wanted a full color display rather than the red on black or black on red provided by the prototype RSD.
Prior studies have either examined a device similar to the RSD called the Virtual Retinal Display (VRD(tm)) without reference to other reading devices or, in the Kleweno, et al. study, its use in comparison to a computer display. In the Kleweno study the VRD was positioned in front of each subject's eye using a chin rest. The average reading speeds obtained with the VRD were comparable to those obtained with text presented on a CRT. In the present study the RSD was positioned using a head band, and this tended to allow the device over time to slip out of position requiring the subject to reposition the device. The use of the chin rest, as opposed to the head mount, may have allowed faster reading speeds, while our study using the head mount examines one clinical configuration of the device.
Our data suggests that in its present form the prototype RSD provides most subject's with reading performance greater than optical devices, but not at the same level as provided by a VM. There was, however, variability in between subject performance that is important to note. In particular, the RSD allowed relatively greater reading performance for subjects with the worst vision, and in two cases allowed subjects to read who could not do so with conventional devices. This finding, thought to be related to the brightness and clarity of the RSD image, is consistent with earlier research. Thus, the use of RSD technology warrants further research to determine its possible role in allowing severely visually impaired subjects to retain functional vision for tasks such as reading.
The application of RSD technology as a low vision reading device is new, however other RSD applications have been in use for nearly two decades. Timberlake and his colleagues , for example, have used related technology in the Scanning Laser Ophthalmoscope (SLO) to study various aspects of functional vision in age-related macular degeneration. More recently Schuchard  and colleagues have begun to examine the use of the SLO in low vision training. Combining the RSD with eye tracking capability would be another approach which would allow us to expand the potential of this technology into the training arena, particularly as the RSD manufacturer, Microvision, Inc. has developed a full color version of the RSD.
In summary, the current study has shown the RSD to have potential in allowing reading by low vision patients for whom conventional reading devices are ineffective. The study has also highlighted a number of areas of potential improvement in the design of the RSD as a reading device. Combined with findings from other research, the results of the present study supports the continued exploration and development of this exciting new technology. This will be particularly true as Microvision continues development of a full-color RSD that focus on ergonomics and feedback from user studies.
This research was supported by a grant from Telesensory, Inc., Sunnyvale, California and the VA Palo Alto Health Care System, Palo Alto, CA. Retinal Scanning Display is a trademark of Microvision, Inc. of Bothell, Washington, the developer of the RSD. Microvision supplied the prototype used in this study. VRD is a trademark of the University of Washington.
1. Goodrich, G.L. and I.L. Bailey, A history of the field of vision rehabilitation from the perspective of low vision, in The Lighthouse handbook on vision impairment and vision rehabilitation, B. Silverstone, et al., Editors. 2000, Oxford University Press: New York. p. 675-715.
2. Kleweno, C.P., et al., The virtual retinal display as a low-vision computer interface: Pilot study. Journal of Rehabilitation Research and Development, 2001. 38(4): p. 431-41.
3. Viirre, E., et al., The virtual retinal display: A new technology for virtual reality and augmented vision in medicine. Studies in Health Technology and Informatics, 1998. 50: p. 252-7.
4. Timberlake, G.T., et al., Retinal localization of scotomata by scanning laser ophthalmoscopy. Investigative Ophthalmology & Visual Science, 1982. 22: p. 91-7.
5. Schuchard, R.A. Using the scanning laser ophthalmosope to assess PRL abilities and characteristics. in Vision '99: Vision rehabilitation: Assessment, intervention and outcomes. 2000. New York: Swets & Zeitlinger Publishers b.v.
Go to previous article
Go to next article
Return to 2002 Table of Contents
Return to Table of Proceedings