Manufacturers Radiation Oncology Medical Physics Diagnostic Imaging Radiation Safety

What's Imporant in PET:
Home  :  Positron
PET technology was introduced in 1975, and the approved clinical applications arrived slowly over the next 25 years. As a technology, PET is mature and the imaging physics are well understood among the developers. PET scanners have changed little over the last 10 years, with the major innovations being an expanding set of clinical applications software and other vendor enhancements that affect throughput. Not unlike other clinical diagnostic products, the availability of reimbursement, combined with clinical need, has caused a rise in the acquisition of PET scanners for clinical applications. Each PET scanner vendor supplies a quality built product. However, the performance of these scanners varies substantially due to each scannerís design characteristics. Some of the important considerations for PET scanning are addressed here. Our goal is to provide sufficient information so that the buyer can independently evaluate the performance of PET scanners without relying solely on the marketing material provided by a particular vendor. Within this presentation we will discuss how Positron Corporationís mPowerTM whole body PET scanner relates to these important design criteria.
Sensitivity is Critical:
  • Sensitivity is critical to accuracy of PET imaging.
  • Localization of lesions in oncology and dynamic quantitative measurement of blood flow in cardiology depend on high sensitivity and high count-rate capability in 2D.
Sensitivity is critical to accuracy of imaging. The more counts obtained the more accurately one can define the locale of the lesion. Localization of lesions in oncology and dynamic quantitative measurement of blood flow in cardiology depend on high sensitivity. In clinical PET, there is a limit to the amount of time allowed to collect data as well as the amount of activity one can administer to the patient. Therefore, every decision made in PET design must optimize sensitivity. The issues surrounding sensitivity are size of detector, distance from patient, dead time, detector encoding scheme, electronics, and packing fraction. The type of crystal being used may impact a particular detector design, but in general it is the efficiency of the detector that maximizes sensitivity.
mPower Whole Body PET
  • Highest sensitivity and count rate in 2D.
  • More uniform slice-to-slice sensitivity.
The sensitivity of Positron Corporation's mPower whole body PET scanner is the highest in the industry. The packing fraction of the detectors is key to sensitivity. The more crystal mass placed around the object being imaged, and the fewer gaps for housing or reflector material, the better. mPowerís unique shorter septa is a result of our patented staggered crystal design. The staggered crystal design allows us to cut the septa length in half as compared to other PET systems and provides twice as many elements, to produce the same solid angle. This allows the detectors to be closer to the subject, increasing sensitivity without decreasing the aperture size. Being close to the patient requires that the detector be capable of accepting and processing higher fluxes of events as compared to other PET scanners for the same amount of activity in the patient. To do this the detector must be smaller and use a simple encoding scheme.
  • The more uniform the sensitivity, the more uniform the quantitation from slice-to-slice.
  • Collecting counts with the septa out (3D) results in a higher level of noise, randoms, scatter, and degrades clinical resolution resulting in reduced accuracy.
Decreasing the size of the septa for increased sensitivity produces another benefit when combined with the mPower staggered crystal design: uniform slice-to-slice sensitivity. The more uniform the sensitivity, the more uniform the quantitation from slice-to-slice. Other designs take an "All or Nothing" approach. Data is collected without a septa for greater sensitivity, or data is collected using a classic septa that has much longer elements, producing a small solid angle and limiting the higher order cross-plane information.
Clinical Resolution & Sampling
  • Asynchronous Coincidence Processing (ACP) increases sampling and maximizes clinical resolution.
Although the size of Positron's Bismuth Germanium Oxide (BGO) crystals is slightly larger than that of other systems, using our exclusive ACP technology we are able to achieve the best clinical resolution. Clinical resolution is a function of detector size, sampling, and sensitivity. In a clinical setting one must have sufficient counts in order to achieve the resolution potential that is determined by the crystal size. Proper sampling allows the clinical resolution to approach the intrinsic resolution determined by the crystal size. Positronís exclusive ACP technology improves in-plane sampling and translates the staggered modules into the same positions as the alternate (non-staggered) modules to more completely sample all slices. Positron's crystals are only slightly larger in the axial direction than they are in the in-plane direction. This results in resolution that is essentially equal in all three directions, eliminating partial volume concerns that exist in other system designs.
PET Detectors
  • Persistence rate is key to sensitivity, count rate capability, and detection accuracy.
  • 4 crystals to 1 photomultiplier tube.
Each PET scanner vendor uses a different detector design resulting in different performance. Detectors designed with larger blocks of crystal and fewer PMTís have reduced sensitivity and count rate capability. In addition, designs using block detectors experience a persistence problem. As the detector is exposed to the typically high count rates experienced in clinical PET, the time to recover after being hit by a gamma (the persistence rate) is critical to the count-rate capability, and ultimately the clinical resolution of the scanner. As gammas hit the block detector, and there is a delay in counting due to the persistence rate, there is a tendency for the detector electronics to misinterpret the location of the event. With high persistence, the signal tends to migrate toward the center of the block. This results in a reduction in accuracy for lesion detection. To minimize this problem, Positron employs a 4 crystal to 1 photomultiplier tube detector design.
PET Detectors
  • Detectors operating in 3D reduce uniformity and accuracy.
To improve sensitivity, competitive designs operate without septa or in 3D mode. When a scanner is operated without septa, the scatter component of all slices is increased. Since the slice sensitivity profile of a wide-open septaless system is triangular, the sensitivity improvement is not the same for each slice. The central slice has the highest sensitivity and the two slices at the extremes of the FOV have the lowest sensitivity. As a result, slices near each edge of the FOV have a degraded trues-to-scatter performance. Also, the slice-to-slice quantitation follows sensitivity and is not uniform, which is undesirable. Competitive designs reduce this effect on quantitation by limiting the acceptance of the highest order cross-plane coincidences, effectively chopping off the top of the triangle reducing any improvement in sensitivity. Positronís mPower achieves high sensitivity, uniformity, and accuracy with septa in place.
  • The key to the best PET scanning is a simple and efficient detector.
Positron utilizes a sub-module-based encoding scheme employing a unique patented one-dimensional light-guide that matches eight BGO crystals (two columns of four) with two PMT's, for a 4:1 crystal-to-PMT ratio. This basic detection unit is much smaller than that of all other competitive designs, minimizing the probability and the effect of secondary gamma detection while handling a previous event. A full module is composed of four (4) such sub-modules, making a full system to be composed of five hundred twelve (512) sub-modules. The detection electronics are highly parallelized to minimize electronics dead time.
  • Volumetric Sampling. Meeting the Nyquist Criteria.
Positron's PET systems truly perform proper volumetric sampling. In order to properly sample the patient with data from a single scan, Positron staggers every other module one-half crystal separation in the axial direction. This provides twice the number of slices in the same volume and satisfies the classic Nyquist sampling requirements to properly capture data from an object. (In this context the Nyquist sampling rules state that in order to fully capture an object, the separation between samples must be equal to, or smaller than one-half the reconstructable resolution of the object). The axial stagger combined with the septa design and use of higher-order cross-plane coincidences, produces a volumetric capture of data that provides extremely uniform quantitation over the majority of the axial FOV. In addition, since the septa do not need to be removed in order to achieve higher sensitivity, the scatter performance of the system is not compromised.
  • The importance of a large field-of-view.
  • Achieving uniform quantitation at the scan boundaries.
Positron mPower axial FOV is 16.6 cm. This allows the system to capture the whole brain, or the whole heart, with a single scan. In addition, such a large axial FOV with uniform slice-to-slice sensitivity facilitates whole body sequential scanning with a minimal three-slice overlap at the edge of each sample. This overlap adds the sensitivities of the end slices from each successive scan to obtain a slice sensitivity that is equal to the central slices. Therefore, this overlap produces uniform quantitation at the scan boundaries.

Harpell Associates Inc.,
1272 Speers Road, Unit 2, Oakville, Ontario, Canada, L6L 2X4
Phone (905) 825-2588, Fax (905) 825-0234, Toll Free: 1-800-387-7168