Today MSNBC.com published an article by Eric Bland for Discovery about the Army developing 'synthetic telepathy' for sending messages.(http://www.msnbc.msn.com/id/27162401/). The article is interesting and thought provoking, which may be a problem if synthetic telepathy becomes real. Mike D'Zmura of University of California at Irvine is the leading researcher for this Army project and Paul Sajda of Columbia University is also working on similar projects. Both scientists talk about the potential usefulness of such equipment and the challenges in making it really possible. Applications as diverse as controlling video games and helping sufferers of Lou Gehrig's disease communicate were mentioned.
Some fancy studies are being done to evaluate feasibility. The three main methods used are functional magnetic resonance imaging (fMRI), magetnoencephalography (MEG), and electroencephalopathy (EEG). Bland writes,
They compare it to current software for dictation systems.
The map generated by all three technologies will help the computer guess which word of phrase a person means when a part of the brain is lights up on the EEG.
The keyword here is "guess."
If anyone can make this real, Michael D'Zmura, Ph.D. can. His curriculum vitae is impressive (http://www.socsci.uci.edu/~mdzmura/). This is no snake-oil salesman. The keyword here is "if."
Likewise, if anyone can find useful medical applications for the work of D'Zmura, Paul Sajda, Ph.D. can. (http://liinc.bme.columbia.edu/mainTemplate.htm?liinc_projects.htm). Another researcher with amazing knowledge and experience.
Dictation systems for medical application were rumored to be in the works about 20 years ago. We who worked in medical transcription knew it would not work. And we were right--then. Now is a different story. Many hospitals, including one of the employers of this writer, either do or shortly will have working dictation systems that transcribe reports directly from speech. Technology has improved enough to make that possible now. There were huge obstacles. For one thing, mumbled speech is practiced diligently by almost all physicians. Foreign-born physicians who mumble are even more difficult to understand. Physicians make up their own words. "Boginates" equals "boggy turbinates" which is a nasal sinus condition that you will not find in any medical dictionary. Variable meanings take their toll. A "tick" that bites and spreads illness such as Lyme disease or Rocky Mountain Spotted fever is not the "tic" of tic doloreux nor the "tic" of diverticulosis. Physicians use abbreviations. When dictated, "BNP" sounds like "BMP" and "CNS" sounds like "C&S." The seasoned transcriptionist knows to anticipate when each of those terms is likely to occur. Medications can be confusing. Way too often "Xanax" sounds like "Zantac" and with a mumbler, "digitoxin" can easily sound like "digoxin." The transcriptionist knows the usage and dosage and just knows which should be the correct term. Technology is here that can learn how doctors speak and guess correctly (eventually) most of the time and fix things that turned out to be correct. This new technology can do things such as put 'BNP or BMP' in a holding pattern until either one result (for a BNP) or several results (for BMP) have been dictated, and then put in the correct spelling. The new technology solves the problems of messy dictation the same way real live well-educated, long-experienced transcriptionists do: it learns by experience to anticipate the probable but has flexibility enough to accommodate surprises. Did I mention that the vocabulary for physicians and medical transcriptionists could easily be well up to a half million words (124,000 medical terms, 225,000 regular words, thousands of medication names, chemistry terms, acronyms, and brand names of equipment). Then there was the matter of background noise. During physician dictation the transcriptionist can also hear conversations between nurses with intimate details that are none of our business, patients wailing or moaning in pain, equipment clanging and thumping, and fire alarms. The new equipment has special phones that minimize the background noise. The old method by dinosaur transcriptionists was to simply ignore it.
This new dictation system will be wonderful for the repetitive things. A normal chest x-ray is a normal chest x-ray is a normal chest x-ray, for example. Many physicians already have "canned" reports for their common dictations. If a total knee arthroplasty is exactly the same all the time, except right versus left, all the physician needs to tell us is which report to use and whether to insert left or right. Most of the canned reports have a few more variables than that, but you get the idea. If the physician can tell a transcriptionist person to do that, they can also tell a machine to do it and that would save a lot of time and work. Probability plays a role in usefullness of this new technology.
The new dictation equipment will cost my employer the equivalent of pay for ten seasoned human transcriptionists for one year and there will be a yearly service fee equivalent to the yearly income of 2.5 human transcriptionists. The odd thing is, all of our jobs are not in jeopardy because someone has to proofread all of these documents. Someone has to transcribe dictation by the worst dictators whose dictation even the best technology cannot decipher. In other words, the hospital will be paying for 14 human workers plus essentially 12.5 more workers for at least the first year.
"Guess" will continue to be the keyword.
Back to the synthetic telepathy, D'Zmura and Sajda have the vision, knowledge, and creativity to make it real to a point. The limitations to synthetic telepathy will be the same as with voice recognition dictation-transcription equipment: human mumbling. If one person cannot understand another standing face to face, receiving all the verbal and nonverbal clues possible, will technology ever overcome the messy knots of frayed ends of cognitive mumbling? Even so, there are some people who can and do think of the same thing at the same time over vast distances and can almost "feel" what the other person is thinking and feeling--without any funny helmets or computers or electrodes--that defies probability. They just know.
In the next few years technology may, indeed, allow a quadriplegic or an ALS sufferer to "think" the wheelchair to move forward or to stop as well as far more complex activities. Learning to ignore the background noise and speak clearly will be bigger obstacles than technology. No more guessing.
By the way, a bologna amputation is never the same thing as a below-knee amputation.