The Evolution of SPECT
The evolution of SPECT imaging has taken many twists and turns as both equipment technology and new radiopharmaceuticals have been developed over the past decade. PET/CT utilization may be experiencing gradual growth, but SPECT remains uncontested as the most well-established and widely used modality in the molecular imaging space. Molecular Imaging Insight checks in with Mark Madsen, PhD, a professor of radiology, physics and astronomy at the University of Iowa Carver School of Medicine in Iowa City, to get a report on some of the latest developments in SPECT hardware, software and clinical applications.
The past several years have produced exciting innovations on both sides—in hardware and image processing. There have been some developments at the component level that enable systems to be customized for particular applications, such as in cardiology. There are new clinical devices that do not use conventional collimation, and imaging systems that reconstruct images using sophisticated integration of SPECT and CT data. While there has been consistent improvements along the way in reconstruction algorithms, new processing algorithms have become commercially available that provide substantial reductions in SPECT acquisition time without sacrificing diagnostic quality.
From a commercial market perspective, SPECT cameras are ubiquitous in healthcare facilities. Cardiac PET imaging gained some utilization momentum due to the SPECT isotope shortage, but for the most part, SPECT remains the workhorse of cardiac imaging. According to the Global Nuclear Imaging Systems / Equipment Market (2011 - 2016) - PET (Positron Emission Tomography) & SPECT (Single Photon Emission Computed Tomography) Landscape report published by markets and markets in 2011, PET is still predominately used in oncology. SPECT, on the other hand, holds a 90 percent share in cardiology. However, PET is catching up with new radiopharmaceuticals, specifically for cardiac PET.
Cardiac SPECT has decades of research behind it and most cardiologists are comfortable with the technology. With respect to cost, a SPECT camera is still much less expensive than a PET scanner.
By the numbers
In terms of healthcare financial burden, heart disease costs the U.S. $108.9 billion each year, according to the most recent National Vital Statistics Report (2013). One in every four deaths in the U.S. is attributed to heart disease; a mortality rate of 600,000 people per year. Heart disease remains at the top of the list as the leading cause of death for both men and women, according to the Centers for Disease Control and Prevention (CDC).
To understand the need for cardiac SPECT, it is important to consider the historic utilization of SPECT as a proven visualization tool for myocardial perfusion, as well as the consistently high and growing prevalence of heart disease (Circulation. 2011;123:933-44. Epub 2011 Jan 24).
This total includes the cost of health care services, medications, and lost productivity. We also know from 2011 data provided by the American Heart Association, that an estimated 7.3 million Americans who suffer from heart disease are uninsured. (AHA, August, 2011).
Logic dictates that considering the cost of the equipment itself, and the reimbursement for procedures, as well as the demographics of the populations in need, SPECT will continue to be cardiology’s mainstay despite the small niche PET/CT has carved for delivering perfusion information and coronary artery calcium scoring.
“Sixty percent, and maybe more than that, of nuclear medicine studies are for cardiology,” says Madsen. There are eight million myocardial perfusion studies done a year. It’s not likely that PET will catch up to that. PET studies, using Rubidium-82, are very expensive. Several new cardiac PET tracers are in development, so they may offer less expensive alternatives, but only to Rubidium studies. The SPECT studies are, by far, less costly and reimbursed at a higher rate.”
Implementation path
Continuous improvements in image quality will go a long way when supporting that great a need. A little more than a decade ago, CT was introduced to SPECT imaging, following in the footsteps of PET and PET/CT. As on the PET side, the addition of the CT for localization purposes has had a tremendous effect on diagnosis and treatment of disease. On the surface, it could be perceived that the improvements in the systems have come mostly on the CT side, going from the initial non-diagnostic model to today’s sophisticated diagnostic multi-slice CTs. Madsen agrees that there have been contributions from the CT side, but it definitely hasn’t done all the heavy lifting in terms of innovation.
“Since SPECT/CT was introduced, there have been a number of changes over the years,” says Madsen. “Initially, the hybrid modality was introduced by adding an additional non-diagnostic CT to the SPECT studies just for the coregistration and the systems weren’t very well integrated together. That’s been continually improving as time goes by. Each vendor has taken its own approach in terms of the types of instrumentation they’ve provided.”
From a hardware perspective, the first entry into the market was a very low power non-diagnostic CT that was added to the SPECT system. Other models soon followed suit, but with diagnostic quality CTs. SPECT was next updated with flat panel detectors, which signified a revolution in hardware and performance. The past two years have been something of an age of enlightenment for SPECT with seamless integration of data reconstruction between modalities , reports Madsen.
Other hybrid SPECT systems such as SPECT/MR are just making their way into the industry. It will likely take a few years to glean how these systems compare to PET/CT and PET/MR and what specific clinical applications are best imaged this way.
Why fix what’s not broken?
While there has been some innovation to the SPECT camera hardware components, the foundational design of the camera itself hasn’t changed much, except in a few cases, according to Madsen.
“We’re not using the [original SPECT camera’s] 60 year-old design. There’s a few new designs that are complete departures.”
Specifically for cardiac imaging, there have been innovations to the actual design of the camera. No longer based on scintillation, the cameras are made using solid state, Cadmium Zinc Telluride (CZT) detectors. CZT is a semiconductor that directly converts x-ray or gamma-ray photons into electrons.
The new solid-state cameras operate in a completely new way. There is no scintillation at all. The CZT absorbs the radiation and directly turns it into an electronic signal. It operates at room temperature and can process more than 10 million photons per square millimeter. The unique combination of spectroscopy and very high count rate capability at room temperature makes CZT an ideal detector solution for SPECT applications.
“It’s a pixilated detector,” Madsen explains, “so the resolution is determined by the location of the detector that gets hit.” Unlike older camera technologies that average signals to determine the approximate event location. Whereas CZT detectors create a geometric configuration of events.
“There are at least two cardiac SPECT systems that have been on the market for three or four years that are based on CZT solid state detectors,” Madsen says. One of these systems is a dedicated cardiac SPECT system without a CT component. Other options exist if hybrid SPECT/CT is the ideal configuration.
The other differentiating factor with these two systems is the collimation. Some top-of-the-line systems use a set of 19 pinholes to provide the imaging aperture rather than a conventional parallel multi-pinhole collimator.
“The collimation changes do affect image quality,” says Madsen. “It’s mainly a difference in approach. With parallel hole collimators, the object was always to sample the entire body, even though you only cared about the heart [in cardiac imaging], but in order to get accurate attenuation and in order to make sure you’re not getting undo influences from activity outside the heart, the thought was always that you would collect all the data.”
The newest systems focus pinholes right on the heart and, as a result, the limited field of view (FOV) achieves a much higher count sensitivity than can be obtained using a regular parallel hole collimator. Other systems have no moving parts at all and have 19 pinholes that project 19 different views of the heart onto this solid state detector. “The trick is to position the patient so the heart is fully visible in each view and there is no truncation,” counsels Madsen.
In terms of workflow innovation, solid-state detector systems can acquire a myocardial perfusion study in just 2-4 minutes. Acquisition of image data from conventional rotating gamma camera systems typically takes on the order of 10-15 minutes. “There’s a big gain in the relative count sensitivity in the two systems over conventional rotating scintillation systems,” he adds.
Some designs includes nine individual mini gamma cameras inside the gantry and each one has a parallel hole collimator that swivels back and forth. A scout study is completed to determine the heart’s approximate location, and the mini cameras swivel back and forth and sample a fairly large area. Because of that type of sampling, it results in a very large increase count sensitivity.
Still other systems have specially designed collimators convergent in the center of the FOV and then parallel toward the edges, a configuration which makes it possible to achieve a higher count sensitivity and magnification for objects that are falling on the center of the FOV of the collimator.
Reconstruction
Over the past five to seven years, a number of commercial products were introduced that compensated for the blurring often seen in nuclear imaging. This resolution recovery increases the apparent count density of the image and improves the noise characteristics so the processed image looks better. These improvements have allowed for shorter acquisition times. It has also provided physicians with an opportunity to reduce the radiation dose to the patient (while imaging time remains longer). Most major vendors have introduced a proprietary version of resolution recovery software.
“My own point of view is that it’s better to reduce the imaging time, rather than decrease the dose,” offers Madsen. “That’s simply because the risk of radiation induced cancer to the age group that’s typically getting myocardial perfusion scans is pretty low, but people die all the time from cardiac disease so getting a good quality scan is important. There’s no question that the quality of the image will be better if people only have to lie still for 2-4 minutes, rather than having to lie still for 10-15,” he says. “I don’t know that everyone would agree with me. There are two sides to the argument.”
True Integration
Innovations in reconstruction software that stand out provide seamless integration of data from both SPECT and CT, says Madsen. “The images they show are very impressive in terms of the improvement that you get when you integrate the datasets in the reconstruction process, instead of after the fact. The CT data are actually part of the reconstruction. It allows them to come up with quantitative measures as well, much like the standard uptake value in PET.”
Future innovations
What is the path to better quantification? According to Madsen, that’s where we should be headed. Superior algorithms could do a better job of incorporating CT.
“The main thing is to achieve a very accurate attenuation correction which also requires that you have a pretty accurate correction for scatter from sources you’re trying to quantify,” he says. “You want to have a system that accurately corrects for attenuation, scatter and resolution losses, and then you want to couple that with some way of validating what the actual activity is and whether that would require the imaging of some standard sources or not, that’s typically how we do it in research studies. I think the industry knows where to go and is already starting to develop these tools.”