Home

Cover Story

Table of Contents

E-Newsletter

Article Archive

Editorial Calendar

Datebook

Writers' Guidelines

Orgs/Links

Opinion Polls

Reprints

Forum

Search

For other articles and previous issues click here.

April 12, 2004

Fusion Profusion
PET/CT Leads the Way But Is Not Alone
By Beth W Orenstein

Combining anatomic and functional images just makes sense. Here’s a look at where fusion imaging stands today and where it might be headed.

The first commercial device to combine PET and CT was designed by a collaborative team of the University of Pittsburgh and CPS Innovations, Knoxville, Tenn. The PET/CT device, partially funded by a grant from the National Cancer Institute, was built by CPS Innovations and clinically evaluated at the University of Pittsburgh Medical Center from 1998 to July 2001.

It hasn’t taken those in nuclear medicine and radiology long to realize the benefits of a combined system. Sales prove it.

Five years ago, all PET scanners were PET alone, says Jonathan Frey, director of PET marketing for Siemens Medical Solutions Nuclear Medicine Group, distributors for PET/CT devices manufactured by CPS Innovations, which is a joint-venture of Siemens Medical Solutions and CTI Molecular Imaging, Inc. (CTIMI). Today, sales of PET/CT systems triple or quadruple the number of stand-alone PET systems. What has spurred this interest in and growth of PET/CT systems?

Hardware and Software
The answer, those in the field say, is continuing improvements in both hardware and software to fuse images from multiple modalities, which has resulted in a growing number of diagnostic and therapeutic applications in the treatment of cancer, cardiac disease, and neurological disorders.

An increasing number of studies show PET/CT to be more effective than PET or CT alone in the diagnosis and staging of various forms of cancer, says Ward Digby, PhD, director of marketing and product management for CTIMI, a supplier of products and services for PET.

A study in the June 19, 2003, issue of The New England Journal of Medicine found PET/CT imaging even more accurate than visually correlating PET and CT results, says Karthik Kuppusamy, PhD, MBA, global manager for functional and CT advance applications for GE Healthcare, Milwaukee. In that article, the researchers suggest that once integrated PET/CT becomes more widely available, it will, for example, be the preferred approach for determining the stage of disease in non-small cell lung cancer, Kuppusamy says.

Originally, Digby explains, the PET/CT fusion was accomplished through software—overlaying the separately acquired PET and CT images on a computer. One problem with that approach was that PET scans and CT images were taken at different times, with the patient lying in different positions. “The patient would not have necessarily laid on the scanning bed in the same way, and the internal organs may have shifted,” he says.
Unless the scans were of the skull, which is rigid, the early software couldn’t bring the images obtained from PET and CT into perfect alignment, Digby says.

Within the last few years, however, manufacturers of PET and CT scanners have developed their equipment so the images can be taken sequentially in the same exam and a single image of both metabolic function and anatomic structure can be displayed almost immediately on the workstation. Improvements have also been made in fusion software to compensate for movement, such as respiration that occurs during image acquisition.

Better Images
With the new PET/CT scanners, “the procedures are done back-to-back and the patient lies down [only] once,” Digby says. “Thus the patient is in the exact same position, and it has just made it possible for every patient’s CT and PET scans to be perfectly registered.”
Digby notes that the scans are done sequentially vs. simultaneously because the CT can be done so fast—in less than one minute. It didn’t make sense to try to do them together, he says. A PET scan, with the most up-to-date scanner technology, typically takes 10 minutes to 15 minutes.

Since CTIMI introduced its PET/CT system, called Reveal, at RSNA 2000, it has made several improvements based on marketplace feedback. The system is now in its third generation, Digby says.

CTIMI formally released its latest Reveal PET/CT—the Hi-REZ—in December 2003. David Townsend, PhD, coinventor of the PET/CT, now of the University of Tennessee (which has the first Hi-REZ system installed and in clinical operation) was a keynote speaker.

The Hi-REZ, which includes a 16-slice CT scanner, offers exceptional image quality. “It can detect lesions about three times smaller than the typical PET detector that has been out there for years,” Digby says. Sixteen-slice CT is state-of-the-art, he adds, and physicians who use PET/CT understandably want state-of-the-art in both modalities. The improved image quality with PET/CT devices is the result of advances in detector technology.

The Hi-REZ system uses lutetium oxyorthorsilicate (LSO) crystals, which CTIMI has found to be the best PET scintillator material available, Digby says. “LSO has a better affinity for the gamma rays that we’re trying to detect,” adds Frey.

Patient Comfort
The latest PET/CT scanners from CTIMI are also more patient-friendly with larger openings to minimize patient claustrophobia. The patient lies on a super rigid bed made of carbon fiber that automatically moves every few minutes through the CT and PET scanners, Digby says.

The hole from some vendors used to be tapered when the patient reached the PET side, but as a result of patient and physician feedback, it has been redesigned so that it remains 70 centimeters all the way through, he says.

The disadvantage of PET/CT devices—even the newer models—is that they don’t tilt like dedicated CT systems. Without being able to tilt the patient bed, it makes it more difficult to perform CT-guided biopsies, Digby says. However, he adds, “tilt has faded as an issue over time because the functional information provided by the PET to guide the biopsy more than makes up for the access challenges lost from not tilting.”

GE Healthcare, a pioneer in the commercial introduction of PET/CT scanners worldwide, has also made significant improvements to its PET/CT scanner devices—its Discovery series is in its second generation, Discovery ST. Discovery series systems are in use in the top five cancer centers as identified by US News & World Report in 2003, Kuppusamy says. The balanced design of the Discovery PET/CT system has substantially increased the overall system sensitivity and reduced scan times, he adds.

Since introducing the system in 1998, GE has redesigned its complete PET/CT system, which today is the only state-of-the-art system that is able to acquire meaningful images in two-, three-, and four-dimensional (2-D, 3-D, and 4-D) modes, Kuppusamy says.

2-D and 3-D
GE’s PET/CT scanners use bismuth germinate crystals, which, Kuppusamy says, have the highest system sensitivity in the market today (sensitivity being one of the critical parameters for maximizing the ability to detect lesions).

Another improvement, and one that differentiates GE from the competition, he says, is the ability to do both 2-D and 3-D scanning in 12- to 22-minute exams for all patient sizes with excellent image quality.

Two-dimensional imaging has proven preferable for larger patients (those weighing more than 180 pounds), Kuppusamy says. “The reason is that two-dimensional imaging is more immune to noise or scatter that comes from the patient. And, if you’re more immune to noise, you can do a better quantitative analysis of the data.”

Having the option of doing 2-D or 3-D scans means that physicians don’t have to compromise confidence when scanning larger patients, he says.

Two-dimensional imaging is also considered a gold standard for cardiac-gated imaging, as well as respiratory-gated imaging, Kuppusamy says.

GE’s PET/CT scanners also offer 4-D imaging, the fourth dimension referring to time. Four-dimensional images help solve the problem of motion that occurs when scanning the heart or lungs, Kuppusamy says. “Four-dimensional imaging is comprised of 2-D and 3-D dynamic imaging, 2-D and 3-D cardiac-gated imaging, and 2-D and 3-D respiratory-gated imaging, and with powerful Discovery systems, the physicians can routinely use 2-D, 3-D, and 4-D imaging.”

GE’s commercial PET/CT scanners are available in four-slice, eight-slice, and 16-slice configurations. The improvements in design and technology that have made PET/CT scans better and faster have an economic advantage as well. For example, a center that was doing eight to 10 scans per day on a dedicated PET can now do 15 or more, Digby says.

Software’s Future
While the market seems to be shifting toward combined PET/CT devices, there is still a strong demand for software that combines images acquired separately. “Software fusion still has a big role to play in PET/CT,” Digby says. For example, he says, radiation oncologists may want to compare earlier PET and CT scans to see whether or not a treatment is working. Fusion software is also useful when the PET and CT images were acquired at different sites. Software, such as that available on the Reveal-MVS system from CTIMI, can then compare changes over time, fusing pretreatment and posttreatment studies.

Joel Leong, MD, PhD, executive medical director of Mirada Solutions, Ltd., which was recently purchased by CTI Molecular Imaging, says the competitive nature of the business has perpetuated the notion that hardware and software fusion are mutually exclusive. However, he says, “this [notion] is incorrect.”

Software fusion and hardware fusion “are complementary, and the range of applications beginning to emerge are not possible without this synergistic combination,” Leong says.
Software fusion, he says, can be as simple as laying one image on top of another, but it can also involve complex mathematical algorithms designed to find the best match according to statistical parameters. The latest advances in software fusion, he says, have to do with deformable fusion or nonlinear or elastic fusion. “Deformable fusion, which allows for compensation for movements and changes that are not quantifiable with simple linear movements, is only possible within software,” Leong says.

Flexible Workstations
Another sign of the future of software fusion imaging is that the software is finding its way onto more high-end postprocessing workstations. For example, Vital Images’ 3-D visualization and analysis workstation, Vitrea 2, will soon be available with Fusion7D software through a licensing agreement with Mirada.

At RSNA 2003, CTIMI and Mirada Solutions demonstrated the Reveal-MVS workstation, which is specifically designed for molecular imaging applications with PET and PET/CT.

The system demonstrated at the show included a number of new features and functionalities designed to streamline the clinical workflow. For example, the system features a quick review of cases, advanced DICOM query and retrieve, and a personal workstation that allows clinicians to take cases with them to review at their convenience or to present outside the reading room.

Researchers believe that all these advances in PET/CT devices and fusion software will lead to increased usage. One area they expect the PET/CT devices and fusion software to play a bigger role is in radiation therapy planning.

CT is good at finding tumors but not at showing which areas of that tumor are the most active, says Frey. PET can provide radiation oncologists with that information, and thus they can modulate the dosage accordingly, he says. Also, they can be more precise and avoid dosing normal tissue.

Also, with PET/CT, Frey says, “you can very quickly assess the effectiveness of your treatment.” Previously, patients would have a CT scan to identify a tumor, be treated, and then scanned with CT again.

“With the combination devices, physicians can see whether a treatment is working as they’re giving it and decide immediately whether to continue it or change to another treatment because it hasn’t worked,” Frey says. “This will become a very important area.”

“We have seen increased use of PET/CT by radiation oncologists and radiologists,” Kuppusamy says. “This is because PET provides metabolically active areas of a tumor, complementing the anatomic areas of tumors identified by CT. Having [knowledge of] the metabolically active areas of tumors alters radiation treatment plans in 25% to 30% of the patients.”

Beyond Cancer Applications
Because of the improvements in the fusion technology and scanners, researchers also see a growing role for PET/CT in cardiology and neurology. For cardiovascular applications, “PET provides myocardial perfusion and viability and CT provides high-resolution morphology of coronary arteries, as well as the extent of coronary calcifications,” Kuppusamy says. “Hence, the PET/CT system provides an excellent noninvasive assessment of coronary artery disease.”

Typically, Kuppusamy notes, eight-slice and 16-slice PET/CT systems are recommended for cardiovascular applications. PET is particularly useful in cardiology because one of the unique challenges of imaging the heart is that it is intrinsically moving and changing shape, Frey explains. PET combined with CT allows the cardiologist to see the plaque, determine exactly where it is, and treat it appropriately.

Because PET measures metabolic activity, which may occur before physical changes show up on CT scans, PET/CT may be useful in diagnosing neurological diseases such as Alzheimer’s disease and brain tumors and in evaluating seizures, Kuppusamy says.
However, Digby says, physicians may not rush as quickly to do PET/CT of the brain because it is rigid and abnormalities found on separately acquired PET and CT are easier to locate.

While advances have been made in hardware and software for the fusion of PET and CT, devices that combine PET and MRI are still a concept.

PET/MR: Software Only
“To our knowledge, there are no hardware PET/MR devices available,” Leong says. They don’t exist, Digby explains, because of the technological hurdles of bringing PET devices into the powerful magnetic field needed for MR. Leong says, however, that just because PET/MR hardware devices don’t exist does not mean fusing PET and MR has no clinical value. All it means is that “currently, the only way of achieving this is with software fusion,” he says.

Indeed, he says, “we have at least one clinical site using our [Mirada Solutions Reveal-MVS] workstation for the purpose of breast PET/MR fusion where our deformable fusion is a valuable tool in compensating for breast movement.”

Another clinical site is using PET and MR of the brain for investigation of epilepsy. Fusion allows the correlation of subtle changes on MR with decreased FDG uptake on a PET study. “If we look into the future at the investigation of neurodegenerative disease, PET/MR fusion has a valuable role to play,” Leong says.

Molecular imaging with PET can also be used to recognize Alzheimer’s disease, Leong says. The fusing of PET findings with the presence of changes on MR studies could also indicate the probable success of interventional therapy.

Like PET/CT, PET/MR also has a future role in cardiac investigations, Leong says. A fusion of PET and MR imaging studies could help identify so-called hibernating myocardium and those patients who would benefit from revascularization therapy.

As more fusion hardware and software is introduced to the marketplace and their usage increases, other issues will be debated, the vendors say. For example, physicians will have to determine protocols for when to send patients for PET, CT, and PET/CT or PET/MR.

Reimbursement for PET/CT and PET/MR studies will be another issue that may be slow to change, Digby says. Currently, PET is reimbursable for a number of cancer diagnoses, but the accompanying CT may not be covered by insurance.

— Beth W. Orenstein is a freelance health writer and regular contributor to Radiology Today.

Subscribe to Radiology Today Magazine!