Positron Emission Tomography,

an insight (part 1)

By Etienne Grech

1: Introduction

Positron emission tomography (PET) may be defined as ‘‘a computerised radiographic technique that uses radioactive substances to examine the metabolic activity of various body structures. The patient either inhales or is injected with a metabolically important substance such as glucose, carrying a radioactive element that emits positively charged particles, or positrons. When the positrons combine with electrons normally found in the cells of the body, gamma rays are emitted. The electronic circuitry and computers of the PET scanner detect the gamma rays and construct colour-coded images that indicate the intensity of the metabolic activity throughout the organ involved’’ (Anderson 2002)

2: Historical overview

The concepts behind PET have been around since the 1950’s, however the first PET scanner was developed in 1970. Until recently, its clinical applications were limited due to the lack of advanced hardware and software. Also the cyclotrons, which are used to produce the radionuclides, were quite bulky. Although very expensive, PET scanners have found their way out of research labs and are being used by large healthcare institutions.

3: Principles of PET

The positron is a positively charged beta particle that is emitted from neutron poor radionuclides. As this particle is emitted, it travels several millimetres in tissue depositing its kinetic energy. During this process, the positron interacts with a free electron in tissue and a mutual annihilation occurs. The unique aspect of this conversion is that the resultant conversion is not one gamma ray of 1.02MeV, but two gamma rays each equal to exactly one half that energy i.e. 150 keV. Furthermore, these rays are emitted at 180 degrees to each other. The patient is surrounded with a ring of detectors and electronically coupled opposing detectors to simultaneously identify the pair of photons.

4: The PET camera

The camera used in PET looks like no ordinary nuclear medicine camera. It physically resembles a CT scanner. The camera uses rings of detectors that are either arranged in a circle around the patient, or in a hexagonal array where the detectors are grouped into cassettes. The content of each cassette varies according to the manufacturer, but usually these consist of 256 crystals being viewed by 16 photomultiplier tubes (four blocks composed of 64 crystals and four photomultiplier tubes) with as many as three rings of 16 buckets in a circle (12,288 crystals with 768 photomultiplier tubes). An alternative design to the bucket, is one that uses rings of detectors with 11 crystals in a staggered array being exposed to 6 photomultiplier tubes.

4.1 PET detectors

As in the common gamma camera, PET detectors are composed of crystals and photomultiplier tubes. However positron detectors are usually made of denser material than that used for lower energy single photon detectors. Such materials include bismuth germanate, barium fluoride, sodium iodide and caesium fluoride with bismuth germanate and barium fluoride being used mostly. A common characteristic of these crystals is their ability to emit light when interacting with photons. The light is then directed towards photomultiplier tubes where it is converted to an electrical signal. Unlike other nuclear medicine equipment where one crystal is viewed by one or more photomultiplier tubes, PET detectors are composed of multiple crystals being exposed by one photomultiplier tube. Crystal selection depends upon whether the equipment is to be used in situations requiring a high-count rate or in situations where a high signal to noise ratio (or high image quality) is required. Such a decision should be based upon four basic crystal characteristics. As bismuth germanate and barium fluoride are most commonly used their characteristics will be discussed.

4.1.1 Detector sensitivity

In theory, all PET camera should keep untrue interactions as low as possible so as to be able to process more true interactions. This is a function of detector sensitivity. Sensitivity varies with multiple coincidence detection capability, density discrimination threshold level and detector size. If the density discrimination threshold level is considered to be the same in both detectors, bismuth germanate has an excellent coincidence detection capability (41%) when compared to that of barium fluoride (4%). Such data shows that bismuth germanate has a higher sensitivity when compared to barium fluoride or other detector material.

4.1.2 Random count rate

Random events may be described as photon pairs that reach the detectors at the same time by sheer chance. These may be annihilation photons from two unassociated annihilation events or they may be completely unrelated to the pair of annihilation gamma rays therefore being considered undesirable as spatial resolution is worsened. The higher the threshold discrimination and the narrower the coincidence time window, the fewer random events are detected. Due to its scintillation pulse profile, barium fluoride has excellent characteristic, which reduce the random count rate.

4.1.3 Scatter count rate

If a 511 keV energy spectrum had to be inspected with bismuth germanate and barium fluoride, one would notice that there are more scattered photons due to Compton interaction in barium fluoride than with bismuth germanate. This suggests that the time that is spent rejecting scattered photons would result in rejecting true coincidence counts because of the dead time considerations. With bismuth germanate, there is a lower number of scattered photons to reject and therefore there is more time to accept true coincidence counts.

4.1.4 Light yield

Another important feature in the consideration of the appropriate crystal for the detection of positrons is the amount of light produced per photon interaction. The crystal material with the best light yield is sodium iodide and it had been used in PET for quite sometime for this reason. However, its hydroscopicity and decay times might contraindicate its use.

4.1.5 Scintillation decay time

Scintillation output is usually seen as a rapid rise in the output of light followed by a slower release (decay time). As shown in table 1 bismuth germanate has a very high decay time (300 nsec) when compared to that of barium fluoride (0.8 nsec). This difference in decay time is the main reason why barium fluoride is preferred to bismuth germanate in high-count rate studies.

Table 1 showing the properties of various PET detectors

 4.2: PET vs. SPECT

PET has shown to have various advantages over SPECT (single photon emission computed tomography). Both the sensitivity and spatial resolution of PET are superior to that of SPECT. Furthermore, PET offers the flexibility of labelling biochemical molecules. However PET scanners and cyclotrons are very expensive and when compared to a normal gamma camera and generator, they are considerably large.

 5: Radionuclides used in PET

5.1 Production of radionuclides

Production of positron radionuclides requires a cyclotron. This is a device, which accelerates charged particles in a spiral fashion to high energies by means of an alternating electric field placed between electrodes in a magnetic field.

5.2 Radionuclides

The analysis of metabolic function is essential for diagnosis and therapy. This can be achieved with PET, as the short-lived radionuclides used are attached to substrates so that a biochemical process can be investigated.

Amongst the isotopes of organic elements, only a few radionuclides exist which have suitable properties for in-vivo use. These include oxygen 15 (half-life =2 min.), nitrogen 13 (half-life = 10 min.) and carbon 11 (half-life = 20 min).

If metabolic substrates are to be labelled indirectly, other radionuclides are used. These include: Bromine 75 and 76, Iodine 124 or Rubidium 82, Gallium 68 and Copper 62. One should note that the choice of nuclide depends upon the problem being investigated.