"Most of us take seeing the television for granted," says Dr. Eli Peli, principal investigator of the study, senior scientist at SERI and a professor at Harvard Medical School. "But for the visually impaired, it is very difficult. This is a source of great frustration and discouragement since so much news and entertainment come from tuning into the 'tube'."
As a low vision rehabilitation expert, Peli sees hundreds of patients suffering from vision impairments caused by diseases such as age-related macular degeneration (AMD), diabetic retinopathy and other diseases that impair the central vision. AMD destroys the tiny central part of the retina called the macula. This makes activities such as reading, driving, and watching television extremely difficult. Peli, an electrical engineer and an optometrist by training, has devoted his career to creating and evaluating new technologies to help low-vision patients regain their ability to do these tasks.
The goal of the current OPO study was to determine if people with impaired vision benefited from an individually tuned contrast enhancement of their TV.
Peli and his colleagues used an image-processing device developed for them by DigiVision Inc. that allows them to manipulate, in real time, the contrast of different sized details (or edges) in the video screen to their individual liking. "This is a very flexible device but also quite expensive and complex, and thus could not be brought to market easily," says Peli. Using this device and similar approaches, Peli and his team learned that certain details in an image that are too small will not be visible to someone with central vision loss even with the highest contrast. They also learned --through patients' responses-- that the large details in an image were fully visible--even to visually impaired persons--without modification. "What we were interested in were the image details that could not be seen until we increased their contrast," says Peli.
The current study had two parts. In the first part, 46 subjects with central vision loss were shown a series of still images and asked to find -- by moving a computer mouse on the table--a setting where the image was most visible. Moving the mouse in one direction changed the size of the detail to be enhanced. Movements in the other direction changed the amount of contrast enhancement applied to those details. . Peli and his colleagues found that the patients did have individual preferences. Each patient repeatedly selected a preferred size and a specific level of contrast enhancement. They also found that patients who selected smaller details to be enhanced selected less contrast enhancement. In the second part of the study, 20 of these patients were asked to watch a video from PBS programming and to rank the quality of the moving images every 10 seconds by moving a mouse up and down on a table in front of them. Every 10 seconds, the processing of the video changed, ranging from un-enhanced, to purposely blurred, to enhanced according to the subject's individual preference. Additionally some images were viewed at enhancement levels arbitrarily set to 4 different levels: smaller details enhanced with higher and with lowers contrast than the individual selection and larger details also at higher and at lower contrast.
Peli and his team found that patients did like the images that reflected their own individual settings for contrast and details better than the un-enhanced video. However, personal preferences were only slightly higher for those individual settings than for the arbitrarily enhanced images.
According to Peli, these results mean that enhancement levels could be selected that will be acceptable to most people, simplifying the type of device needed. Peli envisions a small box that could be attached to any television and would offer only a few enhancement settings, making it less expansive and less complicated to use. Belkin, an international technology company, has recently released a video enhancement cable product, RazorVision, which enhances video for normal sighted TV viewers. *DigiVision, the company that developed the Belkin device and holds a patent pending on the technology, is interested in creating another version of the product for visually impaired viewers based on results from this and other image studies by Peli and his colleagues, *(DigiVision, Inc. was awarded a "Connect 2005 Most Innovative New Product Award" for their "Video Enhancement Cable." Connect is a San Diego organization which has recognized innovative new products in the San Diego region for the last 18 years.)
Though not part of the study results, Peli says that people with normal sight who have been exposed to these enhanced videos do not see them as interfering with their enjoyment of viewing. "In fact, they usually don't seem to be aware that the image has been enhanced," he says. "This means that the visually impaired and their family and friends with normal sight could enjoy watching television together.
To view a split screen image that compares an enhanced and an un-enhanced image, go to http://www.eri.harvard.edu/faculty/peli/projects/hopeinsight.html
For copies of the study, please contact pjacobs12@comcast.net.
Schepens Eye Research Institute is a major affiliate of Harvard Medical School and the largest independent eye research institute in the world.
Journal
Ophthalmic and Physiological Optics