POSITRON EMISSION TOMOGRAPHY (PET) IS VERY GOOD AT CREATING three-dimensional images of tissues, and it can detect changes in tissue behavior better than computed tomography (CT) or magnetic resonance imaging (MRI). Neurologists use PET to map fine alterations in the brain, and oncologists use it to measure the spread of cancer and treatment effects. Yet PET is far less commonly used than other imaging technologies, in part because PET machines are more expensive and are useful for only a small array of clinical problems.

But a handful of engineers are seeking to create nimbler, more versatile PET scanners that would allow for a broader range of possible uses and provide quicker, more revealing images.

Japan’s National Institute of Radiological Sciences, for instance, has been working on designs for a PET scanner that could operate inside an MRI—this would not only help save valuable space but might eventually reduce PET scanners’ exorbitant costs. Yet while a few such combination machines have recently become commercially available, they are more expensive than the combined cost of purchasing separate PET and MRI scanners, says professor and imaging physics team leader Taiga Yamaya at Chiba University—and the resolution of the PET images produced by the machines is about the same as standalone PET scanners. Yamaya’s own prototype of a combined machine, in contrast, has a higher sensitivity than other two-in-one models and could have a price tag of 70% less than current models. Another major advantage, Yamaya adds, is that this type of machine would allow multiple biological functions to be observed at once and would reduce the amount of time patients spend being tested.

A radical effort to shrink the cost and expand the use of PET scanners is also under way at the University of California Davis. A PET scan begins with the detectors that pick up signals from radioactive molecules that have been injected into the body. The Davis project aims to break these detectors out of their monolithic housing and make them more flexible, allowing close-up scans of particular parts of the body.

“Our approach is a lot like a Lego kit,” says Ramsey D. Badawi, professor of radiology and biomedical engineering at UC Davis and lead researcher of the team. “You put the pieces together and point them where you want to look.”

The idea came out of Badawi’s work with rheumatoid arthritis patients. His lab had discovered that the flexor tendon of the finger becomes inflamed in psoriatic arthritis. The researchers wanted to examine this symptom in greater detail, but the tendons are so small that they couldn’t be studied without the high resolution of a PET scan. “To be honest, it would have been prohibitive to develop and use a PET scanner just to image the end of someone’s finger,” Badawi says.

The modular tool kit his team developed can be broken apart, reconfigured and used for a number of tasks: to scan fingertips for Badawi team’s own research, to provide detailed images of mice, or to help surgeons make sure that they’ve removed all of the cancerous cells when excising a lump from a patient’s breast. Another advantage of this flexible PET machine is that detectors can be positioned closer to a specific region of interest, allowing them to pick up more information and produce better images, says Badawi.

Such a tool kit could be much less expensive than current PET machines, he adds. And although there are currently no plans for U.S. production, Badawi has received $400,000—half of which comes from the U.S. Agency for International Development—to build a prototype with a team in Egypt, for use there. “In the short term, such a model could hold great promise for helping developing countries to expand their work in biological sciences,” he says. “It will also be a boon for advanced labs.”

Another effort at UC Davis to improve PET has taken an opposite tack: building a scanner with 560,000 signal detectors—10 times more than most conventional machines. This massive scanner, called EXPLORER, has a central scanning unit measuring about six feet long. The machine is fast, collecting as much as 40 times the “signal” of other PET scanners—which means it can perform a test 40 times more quickly or at an injection dosage 40 times lower, or some combination of the two. Clinical trials of the unit are expected to begin this year.

The PET innovations that ultimately take hold will be those that attract commercial partnerships, explains Badawi. Although he and other innovators make some of their breakthroughs available to other institutions, “I can’t ask my Ph.D. students and postdocs to spend lots of time troubleshooting and supporting equipment in other people’s labs,” says Badawi. Better, faster and more precise imaging will come when these inventions become part of the medical mainstream.