Two mathematics professors have developed a new theory that could, if implemented, pave the way for dramatically more precise renditions of digital imaging and audio, according to an article published this week in Exploration, Vanderbilt University's online research journal. In the article, Vanderbilt professor Akram Aldroubi and University of Connecticut professor Karlheinz Grochenig explain their new theory, which they say would improve the quality of various types of digital content, from MP3 and its descendants to medical imaging, by surmounting limitations resulting from present-day methods.
"Our theory -- which is based on a lot of beautiful new mathematics -- can produce more accurate digital representations of all kinds of samples, including those that classical methods handle poorly or cannot handle at all," Aldroubi was quoted as saying. "It generates algorithms [sets of mathematical procedures] that are fast, efficient, stable, and robust."
Laura Behrens, a broadband services analyst for Gartner G2, told NewsFactor that the new theory could lead to advances in the medical industry and have an impact on digital television. Other areas that stand to benefit include video games, video-on-demand and streaming video services.
In order to convert real-world or "analog" data into a digital format, it must be sampled. In other words, if you want to put a drum loop onto a CD, you need to obtain samples of the loop at regular intervals. If you increase the frequency of the intervals, you improve the quality of the sample.
Although digital representations made from today's sampling technologies are generally very accurate, they do not take into account such concerns as deviations in a sampling pattern, the characteristics of a given sampling device, and degrees of often unavoidable noise pollution.
Aldroubi's and Grochenig's new theory takes into account that sampling measurements are not perfect, thus digital representations should be manipulated more efficiently to obtain clearer audio quality.
In the report, Aldroubi, who previously worked at the National Institutes of Health, demonstrated how this new theory radically improves medical imaging. He took three MRI images: the original, a copy in which 50 percent of the data was removed, and a third that was reconstructed from the second using the new algorithms.
The third image was able to return every key element of the first image. The researchers believe that this new technology has the potential to transform the healthcare industry.
Gartner G2's Behrens speculated that the sort of improved medical imaging that could occur as a result of applying these theories to real-world situations would reduce surprises and increase accuracy of diagnosis and treatment.
"For example, a person is diagnosed with a soft-tissue mass in a large muscle," Behrens told NewsFactor. "The MRI indicates it is almost certainly benign, but cannot be removed without also removing the muscle in which it resides.
"The patient prepares for a lifetime of pain management and limitation on activities [when] the doctor orders a biopsy against the remote possibility the mass is cancerous, and lo and behold, the mass is able to be almost entirely removed."
Behrens said that one could easily imagine other scenarios that would have a happy ending as a result of this technology.
Said Behrens: "Leaving aside the acutely real impact on individual lives, the cost implications of more accurate diagnoses [could] run into the millions or billions of dollars."
No More Snow Days
Behrens told NewsFactor that the professors' theory also has the potential to have a real dollar impact on the digital television industry, which has met with a great deal of consumer resistance as a result of sampling issues.
"Consumers, of course, don't complain about 'the sampling issue,' they complain about black squares on their television screens," said Behrens. "TV viewers can and will sit through a weak or snowy broadcast of programs they very much want to watch. But there's no such thing as a snowy digital broadcast. Images freeze and jump on the screen, or show up without significant numbers of pixels, or don't show up at all -- mainly the last."
Behrens said that better sampling could ameliorate this problem and make watching digital TV -- which, by extension, includes video-on-demand and the other streaming video services on which major media companies are pinning at least some of their future hopes -- a much more satisfying user experience.
Bring the Dramamine
Despite the potential for unprecedented levels of photorealism and accurate sound in video games, improved sampling algorithms could have mixed effects on the computer gaming experience, Billy Pidgeon, client technologies analyst for Jupiter Media Metrix, told NewsFactor.
"Greater realism will be effectively applied in sports games but may detract from some fantasy games where hyper-realism will be more effective than mirrored reality," Pidgeon said. "Also, adding greater levels of graphical detail has the initial effect of reducing interactivity, with more attention paid to presentation rather than reaction."
Pidgeon added that another problem gamers face with very realistic first-person perspective games is the possibility of becoming physically ill.
"When movement within the environment mirrors real-world physics too closely, the eye is tricked into believing that the body is moving [while] the inner ear does not receive supportive stimulus," said Pidgeon. "The user can become nauseous" as a result.