Nuclear medicine departments routinely perform diagnostic imaging and non-imaging procedures and, in many hospitals, undertake radionuclide radiotherapy treatments. However, in this chapter, only radionuclide imaging will be described and therapeutic applications are dealt with in Chapter 7. A radionuclide is an unstable isotope of an element that will spontaneously ‘decay’ by the emission of particles and/or electromagnetic radiation. The three most common emissions from radionuclides are alpha and beta particles and gamma rays. There are two types of beta particle, beta minus (β⁻) and beta plus (β⁺), they are both electrons with a negative and positive charge, respectively. A beta plus is referred to as a positron. A gamma ray is a form of electromagnetic radiation and is the only one with sufficient penetrating characteristics to enable it to be detected externally if originating from within the body. Most use is made of a relatively limited number of radionuclides that emit only gamma rays. However, there is a growing use of radionuclides that decay by emitting a positron; this particle is not penetrating but interacts with an electron, both annihilating to form two photons travelling in opposite directions. These photons are penetrating and can be detected as they leave the body. A radionuclide is attached, or labelled, to a pharmaceutical, forming a radiopharmaceutical. This is introduced into the body, most commonly by intravenous injection but also by ingestion and inhalation, and has a known biodistribution in the body. The distribution of the radiopharmaceutical is related to the physiology or function of an organ, tissue or tissue type; this is in contrast to x-ray, MR or ultrasound imaging which, to a large extent, images structure.

Gamma rays are imaged by a device known as a gamma camera. The device used to detect annihilation photons resulting from the decay of a radionuclide emitting positrons is referred to as a PET camera. Gamma cameras are a large area detection device capable of forming an image of the distribution of a radionuclide within the body. The majority of devices have two detector heads, although it is possible to purchase single or triple head equipment. Imaging with a gamma camera can take place with the detector heads in a stationary position, planar mode, or with the detector heads moving around the body through, typically, 180° or 360°. This mode of acquisition is known as single photon emission computerized tomography (SPECT) or, dropping the reference to computerization (SPET). PET cameras work on the basis of a series of rings of stationary detectors, identifying the location of a nucleus decaying via positron emission, by detection of both annihilation photons within a predefined time (nanoseconds). This is sometimes referred to as coincidence imaging.

The most up-to-date gamma cameras and all PET cameras are now manufactured with an x-ray CT imaging device incorporated into the gantry. In both cases, the primary purpose of this ancillary facility is to allow the use of the structural CT (transmission) images to be converted into an attenuation map [2]. This map is used to correct the functional nuclear medicine (emission) images for variation in body tissue thickness and density. Having the CT imaging incorporated into the cameras allows the emission and transmission images to be acquired sequentially in the same patient imaging session. Assuming there is no patient movement, the data derived from the transmission image can be accurately applied to the emission images. This process is applicable to SPECT images acquired on gamma cameras and all images acquired on PET cameras. Consequently, imaging techniques associated with these devices are known as SPECT/CT and PET/CT. Initially, the CT units were not of a standard comparable to the most up-to-date helical multislice devices and the structural images obtained were not of a diagnostic quality. However, gamma cameras and PET scanners can now be purchased with high specification CT units attached and the registration or fusion of high spatial resolution structural images with the sensitive but poor resolution function images results in an imaging device that is extremely effective at identifying and localizing pathology.

Gamma cameras

Gamma cameras (Figure 6.1) are imaging devices used to detect and record the distribution of a radiopharmaceutical within the body. Gamma rays are detected by a large area crystal of sodium iodide doped with trace quantities of thallium (NaI(Tl)), the crystal structure has the property of absorbing the gamma radiation and re-emitting photons of visible light; the crystal is referred to as a scintillator. The crystal dimensions vary but are typically 50 cm × 40 cm with a thickness of about 1 cm. Smaller detector systems are available for specialized purposes, such as cardiac or brain imaging. A schematic of a gamma camera detector head is shown in Figure 6.2. Further information regarding the design of gamma cameras may be found in Cherry et al [3].

Figure 6.1 Gamma camera.

(Bartec Technologies Ltd)

Figure 6.2 Schematic of gamma camera head.

The process of recording a single gamma ray event by the camera is as follows:

• a γ-ray emitted from the body and travelling towards the detector, in a direction approximately parallel to collimator holes (parallel hole collimator), will be able to pass through and interact with the NaI(Tl) scintillator crystal

• the γ-ray will be absorbed by the crystal and the energy will be re-emitted as a plume of photons with a wavelength associated with visible light

• the light photons travel through a waveguide to the photomultiplier tube (PMT) array where they are converted into photo-electrons

• the photo-electrons are multiplied between the dynodes of the PMT, amplifying the signal

• circuitry between the PMTs produces information about the position of the detected event

• the total output from the PMT array gives information about the energy of the detected event. Energy discriminators allow photons of a specific energy, usually only unscattered photons, to contribute to the image.

An image of the distribution of radiopharmaceutical within the body is made up of hundreds of thousands to millions of detected events. Each event is individually detected. The crystal, collimator and photomultiplier array are mounted inside a lead shield. Most gamma cameras will have two detector heads and the orientation of these heads is variable. Two important performance characteristics of a gamma camera, spatial resolution and sensitivity, are, to a large degree, influenced by the choice of collimator.

The function of a gamma camera collimator is to permit gamma rays travelling in a predefined direction reaching the scintillation crystal. This is achieved by absorbing all other gamma rays in septa between the holes of the collimator. A number of different types of collimators are available but, by far the most widely used, are parallel-hole collimators.

Parallel-hole collimators consist of a thick lead plate with several thousand small parallel-sided holes perpendicular to the plane of the plate. For this type of collimator, there is a 1:1 relationship between the size of the distribution of imaged activity and its projection on to the crystal. Thus, the size of the image is independent of the distance from the subject to the detector face. The characteristics of a parallel-hole collimator depend on the hole diameter, hole length and septal thickness (lead between the holes). Sensitivity and spatial resolution are inversely related; the higher the sensitivity the poorer the spatial resolution and vice versa. Parallel-hole collimators are designed to image gamma rays of a specific energy range, usually denoted as low, medium and high energy. Imaging Tc-99m requires a low energy collimator. The size of hole, length of septa and number of holes influence the characteristic of a collimator designed for a specific energy. Low energy collimators are available with descriptions such as high sensitivity, general purpose, high resolution and very high resolution. The exact characteristics of each collimator can vary immensely between manufacturers.

Imaging techniques

Planar imaging

Reference has been made to planar and tomographic imaging. Planar imaging can take a number of forms. The simplest is with the camera directed at the part of the body containing the organ or organs of interest and an acquisition of a predefined number of events or period of time takes place. These are sometimes referred to as spot views. Keeping the camera heads in a fixed position but causing the patient to move slowly between the detectors is commonly referred to as whole body imaging. The bone scans shown in the section on Clinical applications utilize this form of imaging. A dynamic acquisition is when a series of images, often sequential and of the same duration, is obtained. With these data, temporal changes in distribution guide the diagnostic results. An example of dynamic imaging is utilized in the section on kidney imaging. A final form of planar imaging is when the acquisition is synchronized to a physiological signal obtained from the patient. A gated blood pool study shown in the section on cardiac imaging is the most widely used form of this mode of imaging. The cardiac cycle is divided into a fixed number of frames, in the example shown 24 frames each of 41 milliseconds. The dynamic acquisition is triggered by the R-wave on the patient’s ECG. Images from multiple heart beats are obtained and corresponding frames (first, second, etc.) are summed together to produce a dynamic series representing a composite beat. Without this technique, the quantity of data in each frame would be insufficient to achieve the accuracy of the ejection fraction calculation required.

Tomographic imaging

Tomographic imaging is performed with a gamma camera, by rotating the detector system around the patient while acquiring images. The image data may be obtained with the camera heads stationary at a number of fixed positions around the body known as ‘step and shoot’ mode, or acquired continuously as the camera rotates. Two-dimensional (2D) cross-sectional images are reconstructed using mathematical algorithms, in a similar way to x-ray CT. The 2D cross-sectional images can be stacked to form a volume of data which gives the appearance of having acquired a 3D dataset.

Knowledge of the 3D distribution of the radiopharmaceutical within the body allows correction for the attenuation of the emitting gamma radiation if the size and density of surrounding structures can be determined. This is achieved with the aid of a CT attenuation map. A CT acquisition represents the variation of x-ray absorption throughout the image volume, a direct relationship can be made between this and the attenuation of the emitted gamma radiation. An example of nuclear medicine images that have been attenuation corrected and fused with the CT structural images is shown in Figure evolve 6.6 . Note that the CT device in this example is a low dose unit, not in the diagnostic category.

Gamma camera performance characteristics

It is essential that gamma camera performance is maintained to achieve the highest possible sensitivity and specificity of each investigation undertaken on it. The performance of gamma camera detectors is characterized by a number of parameters:

• spatial resolution

• sensitivity

• uniformity

• linearity and energy response

• energy resolution

• multiple window registration.

Detailed descriptions of the precise definitions of these parameters, factors influencing their stability, measurement methodologies and how they may be specified during equipment procurement are covered in two publications; IPEM 2003 [4] and NEMA 2007 [5]. Brief outlines are given here.

Spatial resolution. This is the ability to distinguish an object from its surroundings and is a measure of the sharpness of the image. It is defined by a full width half maximum distance and this is about 3 mm for the intrinsic (no collimator) resolution of a current gamma camera.

Sensitivity. This term relates to the number of gamma events per MBq detected by the gamma camera. Described above is the influence the collimator has on the spatial resolution and sensitivity of a gamma camera.

Uniformity. This refers to the variations in count rate across the field of view when the detector is exposed to a uniform source of a gamma ray emitting radionuclide.

Linearity and energy response. These refer to the gamma camera’s ability to determine the true location and energy of a gamma ray detected anywhere in the field of view. These are the fundamental parameters that influence detector uniformity.

Energy resolution. Statistical variations in the detection of gamma rays by the crystal and photomultiplier assembly result in a characteristic broadening of the total absorption peak of the energy spectrum. Energy resolution is defined by an energy Full Width Half Maximum.

Multiple window registration. This refers to the ability of the gamma camera to determine the location of a detected gamma event as a function of its energy.

Corrections can be made during data acquisition to overcome some of the non-random defects in camera performance. These include corrections for non-uniformity of sensitivity, non-uniformity of energy resolution, and non-linearity.

There are additional performance characteristics to take into account when performing SPECT imaging and still further with SPECT/CT. SPECT imaging requires the centre of rotation (COR) of the rotating camera heads to be within predefined limits. Failure to do so will result in artifacts on the reconstructed SPECT images. It is also important that the parameters defining the detector heads remain stable at every angle of rotation. SPECT/CT relies on the accurate registration of the SPECT and CT images. The performance of the CT is assessed as for all CT equipment, as covered by IPEM Report 91 [6]. A crucial additional assessment is the alignment of the SPECT and CT data together. Phantoms are routinely used to assess the gamma camera SPECT capability and SPECT/CT alignment.

Positron emission tomography scanners

Dedicated positron emission tomography cameras (Figure 6.3) are designed specifically for radionuclides that decay by positron emission. Scintillator crystals with high stopping power are utilized, examples are bismuth germanate (BGO) and lutetium orthosilicate (LSO). These crystals, coupled to photomultiplier tubes, are positioned in rings around a gantry. Electronic circuitry, designed to register if two crystals detect a photon within a ‘coincidence time’, usually a few nanoseconds, allow ‘true’ events to be detected. Many ‘single’ events will be detected. The size of the crystal elements depends on the make and model, but is generally 3 to 6 mm. The elements are arranged in 360° rings, with typically 15 to 30 rings consisting of the order of 10 000 individual crystal elements. The crystal size is one of the main factors in limiting the spatial resolution, commonly 4 mm (FWHM) [7]. Although systems designed for special purposes, such as brain scanning, can achieve 2 mm. The computer system is able to reconstruct the line connecting two detected events in opposite crystals as a line of response (LOR). Millions of LORs are acquired over a period of time and reconstructed to form cross-sectional images using appropriate image reconstruction software.