PET developers focus on ways to cut scan time, boost throughput

August 6, 2003

CPS, Siemens balance speed and diagnostic qualityRadiologists crave speed, but they have never been willing to sacrifice image quality. Executives at CPS Innovations and Siemens Medical Solutions not only realize that, they are

CPS, Siemens balance speed and diagnostic quality

Radiologists crave speed, but they have never been willing to sacrifice image quality. Executives at CPS Innovations and Siemens Medical Solutions not only realize that, they are banking on it. Technological developments in scintillator crystals, specifically the LSO (lutetium oxyorthosilicate) these companies use in their PET scanners, have provided the opportunity to serve both speed and quality, according to executives from both companies.

The proof was clear at the annual meeting of the Society of Nuclear Medicine in June, when an image created using LSO-based PET/CT at the University of California, Los Angeles, was picked as the image of the year. Data for the image were acquired in three minutes, yet the tumor-a metastasis in the neck-was clearly evident.

The achievement validates a primary tenet of CPS and Siemens: that diagnostic images can and should be obtained as quickly as possible. That sentiment was echoed at the SNM meeting by Dr. Henry Wagner Jr., one of the pioneers of nuclear medicine, who chose the image of the year. Speed and image quality need to match the clinical need, he said.

"In many cases, you don't need to know the structure of the metastatic lesion," he said. "You're only interested in its detection."

In his presentation of this image at the meeting, he noted that the current volume of PET procedures is light enough that most practitioners have the luxury of spending as much time as they want to do a scan. But that will change.

"I believe we are moving into a high-throughput world, and we don't want PET and other nuclear medicine procedures left behind because they are rate limiting," said Wagner, who founded the Johns Hopkins nuclear medicine division in 1959.

New clinical applications with short-half-life radiotracers may also demand faster scans. One such tracer is carbon-11 acetate, according to Markus Lusser, vice president of worldwide sales and marketing for the Siemens Nuclear Medicine Group.

"We will need an ultrafast scanner to get the most from these tracers," he said.

A more immediate issue is patient comfort, which improves as scan times go down. Making the patient more comfortable may even translate into better image quality, as the patient is less prone to fidgeting and the images consequently less vulnerable to motion artifact, Lusser said.

UCLA is already breaking the clinical ground needed to speed up positron imaging. Technology plays a role, but so do commonsense management techniques. Physicians and technologists assess patients' body habitus before finalizing a scan protocol, generally taking more time with heavier patients and less with lighter ones. The time spent handling patients is also optimized, as technologists utilize two rooms to administer positron-emitting nuclides. Total exam time is well under an hour, compared with other facilities that take 90 minutes or more, according to the staff.

But scan time must come down further if PET is to reach the throughput of other modalities such as CT and MR, Wagner said. With patient management optimized, efficiencies must be achieved during data acquisition. CPS, a joint venture of Siemens and CTI Molecular Imaging, is working on this part of the equation.

The company, which supplies PET scanners for sale by Siemens, CTI, and Hitachi Medical, has developed a digital imaging chain that promises to at least double the count-rate performance of the existing line of LSO-based PET/CT systems. Company estimates peg the improvement conferred by Pico-3D at a 70% count-rate increase for low-dose isotopes, such as FDG, and double the capability for short-lived isotopes, such as rubidium.

Before the introduction of Pico-3D this spring, electronics on the LSO-based systems were not the best, according to Ronald Nutt, Ph.D., president of CPS Innovations.

"Our electronics were slightly modified from the BGO (bismuth germanate) tomographs to work with those using LSO," he said. "Although they were improved, they were not optimized. The Pico electronics are made precisely to fit the parameters of LSO."

Pico-3D, which was developed by CPS' sister company Concord Microsystems of Knoxville, TN, digitizes signals from LSO detectors every 500 picoseconds. An ultrafast sampling technique allows better scatter rejection and higher contrast imaging. A five-nanosecond coincidence window further improves the rejection of random counts.

"Pico electronics will take us into the future," Nutt said. "It has time resolution in the order of picoseconds instead of nanoseconds."

This may allow PET scanners to do time-of-flight studies. These were first imagined 30 years ago with the advent of PET but were held back by slow scintillators and inadequate data pipelines. Nutt believes improvements in both areas make TOF a reality, raising the prospect of reducing noise in images by a factor of five or six. The only question is when.

"It's something we will be working on in the future," he said. "Maybe in a couple years we will be talking about it."

CPS is examining this and other possibilities using a research platform that employs multiple detector heads in place of a ring detector. Among them is spiral PET. Unlike the stop-and-shoot method now in place, the patient would move continuously through the gantry, albeit slowly. CPS engineers are currently assessing speeds as slow as 0.33 mm per second. Spiral PET might, however, decrease scan time and increase image quality through improved uniformity, Nutt said.

"Every line of detectors would look at every part of the body as it passes by," he said.

Spiral PET would eliminate the miniscule, but not negligible, motion artifact that can come from a patient table as it repeatedly steps forward, as happens in stop-and-shoot scans.

The multihead R&D platform might also serve as the stepping-stone to truly hybridized imaging. CPS engineers are looking into mounting CT detectors and x-ray sources between the multihead detectors as part of a grant from the National Cancer Institute. Ultimately, gamma and x-ray detection might be combined in dual-purpose panels. The company has received two patents for doing so.

"We haven't done a great deal in that development except some of the basic measurements, but it is clear it can be done," Nutt said. "Eventually, we would like to see all the technology in PET and CT converge: the same electronics, scintillators, processors. But this is really blue sky stuff."