Terry Turpin is a man of indeterminate middle age with tentacles of dark hair defying his pate and a pirate-style black patch over his left eye. In the unlikely event that he wears a tie, it will be the all-purpose loud one that he keeps in his office. Anyone who sits next to him on an airplane must get the instant impression that Turpin is some kind of way-out scientist.
Turpin spent two decades doing something in that deepest dungeon of the federal government, the National Security Agency. Now he is chief scientist for Essex Corp., a small defense contractor headquartered in Columbia.
It was on a plane flight in 1989 that Turpin was struck by the inspiration that he hopes will crown his career, and that Essex Corp. is all but counting on for its future.
Turpin was flying along and thinking about -- what else? -- the computational dynamics of image processing for synthetic aperture radar. Whatever it was he did at NSA had to do with using lasers instead of digital circuitry to make calculations, and at that moment he made an important connection:
Computers take forever to process images from super-advanced radar systems, but a simple beam of light, Turpin theorized, could do it almost instantly.
Seven years later, Turpin and Essex are scrambling to pin down the enormous apparent potential of the resulting patented device, called the ImSyn image processor.
It could open dramatic new doors for doctors peering into the brain as it thinks and the heart as it beats. It could make significant changes in how military and commercial pilots use radar. It could create a revolutionary new microscope. It could hold a whole universe of uses its makers haven't even visualized yet.
Or it could wind up as just a laboratory curiosity.
It's a race against time and imagination for Essex. Company executives hope they've found the Holy Grail of the small defense contractor in the 1990s -- a product so useful that it will protect them against the diminishing military market.
But with ordinary computers getting faster and cheaper every day, some question the long-term value of ImSyn.
"For the average user, what you want is a maximum amount of flexibility. You can do many things with a computer. Their device has some very finite limitations," said one defense industry expert. "If somebody asked me to come invest in this thing I wouldn't. But for every guy like me, you'll probably find a guy to say this is the greatest thing in the world."
Harry Letaw Jr., the president of Essex Corp., would settle for ImSyn being the greatest thing in a few small parts of the world.
His company, like so many others, has struggled in the wake of the Cold War. In 1989, Essex had almost 300 employees, sales of more than $29 million and a net profit of $126,000. It has now sustained a loss every year since 1990, revenues have dropped by half and employment has sunk as low as 135 in 1993. Its stock, trading in the $3 range in 1989, has essentially remained flat, and was trading Friday at $2.375.
The picture is slowly improving for Essex -- employment is back up to about 216 -- and Letaw has weaned the company from 100 percent dependence on government work to about 85 percent. They perform engineering to make computers easier for humans to use, design training systems and devise software to measure human performance.
But the future is more or less being staked on Turpin's gadget, ImSyn.
Letaw and Vice President Matthew Bechta like to show off a mock-up of the device, a black box about the size of a stereo receiver with a glass top so you can see the laser and lenses inside. Three years ago, Bechta jokes, they could have wound up in federal prison for displaying the thing. It was classified top secret.
The government finally declassified the material a couple of years ago and encouraged Essex to shop for commercial applications, Bechta said.
The idea was not only to let Essex find a way to survive, but also to create a product with wider applications that would be cheaper for the military to acquire.
Even so, the first sales of ImSyn are for government work. The company has arrangements with three government contractors to deliver an ImSyn processor for various types of military applications -- including a deal with the Electronic Sensors and Systems division of Northrop Grumman in Linthicum to apply ImSyn to a radar unit.
ImSyn is as offbeat a product as its creator's appearance would suggest. Unlike a conventional computer, which must convert data to coordinates on a rectangular grid before processing, the beam of laser light in ImSyn converts the data directly to electrical impulse.
If a computer is comparable to the cold precision of music on compact disc, the analog function of ImSyn is akin to the natural sound of a live instrument. The mathematical shortcut represented by that difference makes ImSyn enormously faster than a computer on some types of calculations.
In addition, some types of imaging don't easily conform to the rectangular grid a computer needs for processing; for instance, a Magnetic Resonance Imager, or MRI, peering into the body of a hospital patient, prefers to scan in a spiral.
The ImSyn's laser light, broken up by crystals in a way that represents numbers and then beamed into a video camera, can handle spiral scans or any weird shape with ease.
Essex is pushing the technology toward three distinct uses. Theoretically, it could provide lighter, faster, smaller gear for airplane radar -- allowing smaller planes to use more sophisticated gear, and letting technicians view certain images in flight instead of reviewing the data on the ground.
This could help pilots track floods through heavy foliage, for example, or on search and rescue missions.
ImSyn also seems to have a natural affinity for the work done by MRI scanners. Moriel NessAiver, a physicist at the University of Maryland Medical School, has been working with Essex under a $75,000 state grant to explore the possibilities.
Where a normal computer may take four hours to process a particular type of MRI image, he said, the ImSyn can do it in less than 10 minutes. And the image quality appears to be much better.
This could mean faster, cheaper and more accurate scans, Ness-Aiver said -- and increased use of the MRI machine.
Even more intriguing is the possibility that ImSyn could be tailor-made for the hottest type of MRI process, Functional MRI. This entails watching bodily functions in action, such as a heart as it beats or the brain as it responds to different types of thoughts and actions.
In April, NessAiver made his first presentation on ImSyn at the annual convention of the International Society for Magnetic Resonance in Medicine. He drew almost rabid response, he said, by describing how ImSyn could process up to 20 images per second.
Later, he said, a team from Stanford presented a paper on its efforts to use the MRI to watch a pumping heart. They corralled a whole network of computer workstations to process the deluge of images.
"And then they said, 'Another option could be the process presented earlier by Dr. NessAiver,' " he said. "So the potential was recognized immediately."
At this point, though, scientific potential is all NessAiver is willing to vouch for. "To be honest with you, I don't know if it will be commercially viable in the MR market. Just because a product is useful doesn't mean it will succeed," he said.
Conventional computers, he pointed out, are doubling in speed every 18 months or so, meaning: "Essex has a limited window of opportunity to get [ImSyn] out there."
The disadvantage the company faces in the MRI market, he said, is the same one it faces in the area of radar imaging: "They're trying to get into an area that already exists."
But, he added, "in the area of microscopy, there's nothing else that does what they're doing."
That's the third field Essex is scrambling to master, and the one currently drawing most of Turpin's attention.
His prototype synthetic aperture microscope doesn't look like much: Shrouded by pieces of cardboard held together with duct tape, the device consists of old camera lenses and other scraps found around the laboratory.
Plugged into an ImSyn, though, the microscope can get three-dimensional images of objects smaller than one micron -- a micron being one-millionth of a meter (the human hair is about 100 microns in diameter).
Most remarkably, the device can get these images from distances of 2, 3 or even 10 inches from the object. The lens of a conventional microscope would have to be in contact with its subject to get such resolution, and then it wouldn't be in three dimensions, said Lee Peachey, a biologist at the University of Pennsylvania who has been advising Essex.
Turpin's new device also does not require dyes to distinguish between different structures in biological samples, Peachey said. "I think it is definitely a significant innovation in microscopy."
The potential for biologists, and for industrial uses such as examining silicon chips on a production line, could be enormous, he said.
Could be -- that's the key phrase. Essex still has significant work to do to confirm that the microscope is as dramatic a step as it appears. And while attention is focused there, the clock is ticking on the race to turn each area of ImSyn's potential into cash.
Pub Date: 9/15/96