Nuclear Medicine Resources Manual

Подождите немного. Документ загружается.

CHAPTER 4. INSTRUMENTATION

130

4.3.3.5. Recommended frequency of quality control tests

The recommended frequency of quality control tests depends on the

particular equipment available and its stability. Thus it is important to tailor

quality control tests to specific equipment. Significant changes consistently

detected between consecutive quality control tests may require the frequency of

the tests to be increased. Conversely, the frequency may be reduced if only

minor fluctuations are detected over a series of quality control tests. Manufac

turers’ literature may also provide some guidance on the required frequency of

tests. An experienced nuclear medicine physicist may in addition provide advice

on the frequency for specific tests and equipment.

Tests such as those on uniformity are specifically designed to detect

malfunction of the equipment and sudden deterioration of performance before

they affect a large number of patient studies. Thus the frequency of this type of

test should not be reduced even if results remain consistent over a prolonged

period of time.

The following schedule is thus recommended:

Daily:

—

Visual inspection;

—Background and/or contamination;

—Photopeak and window setting;

—Uniformity.

Weekly:

—

High count flood;

System dependent frequency:

—

Centre of rotation;

—Energy and uniformity correction;

—Uniformity correction floods for all collimators used for SPECT.

After a major service:

—

Spatial resolution;

—Uniformity with high count flood;

—Multiple window spatial registration;

—Whole body resolution.

4.4. DUAL PHOTON IMAGING

131

Less frequent tests:

—

Pixel size;

—Planar spatial resolution (FWHM);

—SPECT resolution;

—SPECT total performance.

4.4. DUAL PHOTON IMAGING

4.4.1. What is dual photon emission tomography?

In contrast to SPECT, which relies on the detection of single photon

emissions, PET involves the detection of the dual photons that arise from each

positron. When a positron, a positively charged electron, is emitted from a

nucleus, it travels a short distance, losing energy until it reaches a resting state. It

then interacts with one of the many electrons, whereupon the two annihilate

(disappear), giving rise to two 511 keV gamma rays that travel in opposite

directions. Unlike SPECT, where a single photon is emitted on each disinte

gration, in PET a pair of photons is emitted (hence the term dual photon

imaging). Detection involves a pair of opposing detectors, which must record

events at the same instant of time (i.e. in coincidence).

This unique process has two very important properties. Firstly, since the

two photons travel in opposite directions, the point of annihilation will lie on a

straight line joining the points of detection. This means that directional

information is determined electronically, without the need for conventional

collimation. Hence, unlike SPECT, detection is not limited to those photons

travelling at right angles to the detector and consequently the sensitivity of PET

is many times greater than that of SPECT. Collimation may be retained for

separate data from different planes; however, within any detection plane no

conventional collimation exists.

The second important property of dual photon imaging is that for it

attenuation is dependent only on the total attenuating path through the patient,

but is independent of the exact location of the annihilation event in the tissue.

This is quite different from the case of SPECT where attenuation poses a much

more difficult problem.

4.4.2. Dedicated PET instrumentation

Dedicated PET systems consist of multiple rings of detectors that

surround the patient. Each ring normally consists of a set of small BGO

CHAPTER 4. INSTRUMENTATION

132

detectors, the high density of BGO giving it a very effective stopping power for

511 keV photons. Sodium iodide (NaI), the detector of choice for gamma

cameras and SPECT systems, is also used; however, it has several properties

which make it less attractive than BGO. It does have a higher light output and

consequently better energy resolution than BGO (~10% compared with ~20%)

but, besides its significantly lower stopping power, it is hygroscopic.

Dedicated PET systems may use lead collimation in the form of 1 mm x 65

mm lead septa, which significantly reduces scattering and random events.

Recent developments have involved removal of these septa, so that there is no

collimation within the imaging volume of the scanner, necessitating a 3-D recon

struction of the data. This can potentially improve sensitivity by a factor of six.

The differences between PET and SPECT in performance and applica-

tions are decreasing. Dedicated multidetector SPECT instruments with purpose

designed collimators, for example fanbeam collimators, now permit SPECT

studies to be performed with improved sensitivity compared with that of single

head SPECT and comparable resolution to that of PET, although sensitivity is

still lower than that of PET.

4.4.3. Hybrid PET–SPECT instrumentation

Although gamma cameras were used in the early days of PET, manufac-

turers have recently introduced commercial gamma cameras based on

coincidence detection systems. Dual head gamma cameras with opposing heads,

originally designed for multidetector SPECT, can be used with additional

coincidence circuitry to detect positron events in exactly the same way as PET

detectors. The clear advantage of these systems is their relatively low cost and

the fact that the instrument can be used for either SPECT or coincidence

detection (CD).

The term CD is often used to differentiate these systems from dedicated

PET systems, although a more appropriate term is ‘hybrid PET–SPECT

systems’. The only real difference is that the dual detectors must be rotated

around the patient, as in normal SPECT acquisition, whereas dedicated PET

systems are designed with a complete (or partial) ring of detectors that surround

the patient.

The main differences in the design of a hybrid PET–SPECT system,

compared with a dual head SPECT system, are as follows.

Coincidence circuitry must be added so that the two opposing detectors

can detect the two annihilation photons in coincidence, i.e. within a very short

time of each other. It is this coincidence that defines the path along which the

photons must have travelled, eliminating the need for a conventional collimator.

4.4. DUAL PHOTON IMAGING

133

The absence of a collimator means that the sensitivity for detection is

much higher than normal. This introduces problems relating to count rate

performance since, for each detected coincidence, there are many more single

events (single gamma photons detected without a corresponding coincident

event). Several approaches have been implemented to improve the count rate

performance of gamma cameras, with some coincidence systems now capable of

achieving count rates of several million counts per second.

Since for 511 keV photons the stopping power of sodium iodide is

relatively poor, manufacturers offer thicker crystals than normal (up to 25 mm

thick), with only slightly poorer resolution, due to uncertainty in the location of

detected events. The ability to maintain performance is largely attributable to

the improved design of recent gamma cameras. Despite the thicker crystals, the

detection efficiency for 511 keV coincident photons is still relatively low

compared with that for PET systems that use BGO detectors.

It should be recalled that the absence of a collimator means that the

resolution is essentially defined by the intrinsic resolution of the gamma camera

at 511 keV (typically 4.5–5.5 mm).

Dual head systems rotate to different angles around the patient, recording

coincidences at each angle. The detected events are rebinned or sorted in the

same way as in a dedicated PET system to form sinograms for each slice. These

are reconstructed using filtered back-projection or iterative reconstruction in

exactly the same way as in dedicated PET or SPECT. Although the earlier

systems did not include attenuation correction, recent systems now have this as

an option.

Despite the attraction of low cost and the versatility of the hybrid PET–

SPECT systems, their place in clinical practice has still to be defined. The

performance of CD is definitely limited compared with that of dedicated PET

(Table 4.1) and this has imposed limitations in the technique for the detection of

small lesions. Nevertheless, their introduction has resulted in the widespread use

of positron emitting tracers in clinical practice.

4.4.4. PET versus SPECT

It is useful to make a direct comparison of performance between PET and

gamma camera based SPECT, although this largely depends on the choice of

collimator and number of heads. PET performance will also depend on several

parameters, for example detector diameter and type of septa. The figures

provided in Table 4.1

are therefore indicative only. Current PET systems

provide a reconstructed resolution on-axis of around 6 mm. SPECT can achieve

a resolution only slightly worse than PET using ultrahigh resolution fanbeam

collimators (around 8–10 mm), but this can only be achieved with significantly

CHAPTER 4. INSTRUMENTATION

134

less sensitivity: at best a factor of less than 2.5 (with fanbeam and triple head

compared with 2-D PET)

4.4.5. Use of ultrahigh energy collimators

A very simple approach to imaging 511 keV photons is to use an ultrahigh

energy collimator. In this case conventional SPECT or planar imaging is

preferable to coincidence imaging, which is based on the detection of the single

511 keV photons. Although coincidence imaging was used initially in oncology

studies, it has since become evident that only fairly large tumours can be

detected. However, the need for high resolution is less critical in cardiac studies

using FDG since normal or enhanced uptake is evidence of myocardial

viability.

4.4.6. Purchase of dual photon imaging systems

All nuclear medicine physicians, assisted by a nuclear medicine physicist,

acquire some experience during their careers in purchasing gamma cameras

and other accessories for a nuclear medicine service. The decision making

process, leading to the purchase of a system performing dual photon imaging,

calls for knowledge of the basic physics of coincidence detection and of the

differences between 2-D and 3-D acquisition in terms of sensitivity, the ratios

between the true and the random events, and scatter fraction, as well as the

different methods to overcome these problems.

There are a number of different ways to increase the sensitivity of the

system and physicians should work closely with a physicist who has extensive

knowledge of these areas. It is recommended that they should visit or contact a

site that is already functioning. They should be allowed to review the

TABLE 4.1. COMPARISON OF PET VERSUS SPECT SYSTEMS

Characteristic SPECT

Hybrid

PET–SPECT

PET

(2-D)

PET

(3-D)

Sensitivity relative to SPECT ×1 (single high

resolution)

×6 (triple head

fanbeam)

ª×7 ×15 ×75

Reconstructed resolution (mm) 8–10 (head) 5–7 4–6 4–6

Scatter fraction (%) ª30 ª30 ª15 ª40

Attenuation factor (thorax) ×7 (Tc-99m) ×30 ×30 ×30

4.4. DUAL PHOTON IMAGING

135

performance of the system from the work scheduling book in terms of its

uptime and downtime. They should also have an opportunity to observe on the

workstation the studies performed. The nuclear medicine physicist should be

able to review the results of the various quality control tests performed.

There are many aspects of purchasing dual photon imaging systems that

are common to the purchasing of single photon imaging systems; these have

been covered in an earlier section of this chapter. In addition to specific advice

on contractual arrangements, warranty and service, the reader should bear the

following points in mind when purchasing dual photon imaging equipment.

In most cases the primary purpose of purchasing the equipment is to

perform oncology studies, although specific centres may have research require

ments in other areas. The main dilemma facing a purchaser is what type of

system to purchase. At the time of writing there are basically two types of

system: dedicated PET systems (including full ring systems and lower cost

partial ring systems) and hybrid SPECT–PET systems based on standard

gamma camera technology, and there is a significant difference in the cost of

these systems. The main considerations in choosing between the systems can be

summarized as follows.

(a) Sensitivity

In a hybrid PET–SPECT system, sensitivity is a primary concern and a

thicker crystal than that normally used for SPECT is required. The effect of

increasing crystal thickness on routine single photon nuclear medicine studies

should be considered. Although a slight decrease in resolution is demonstrated

in bar phantom studies, it has little effect on routine clinical studies. An

advantage is the additional increase in the sensitivity for such radionuclides as

Ga,

111

In and

131

I. The trade-off between resolution and sensitivity in PET

versus SPECT applications for these thick crystal systems is still under

evaluation. Sensitivity is improved by using 3-D rather than 2-D acquisition as

outlined in the sections earlier in this chapter. The exact trade-off in useful

counts (with scatter correction) for whole body applications continues to be

evaluated.

(b) Count rate

The use of a large single detector results in specific count rate limitations.

There are several approaches to improve count rate capability with specific

circuitry designed to enhance the performance of gamma camera based

systems. Nevertheless, there is a limit to the activity that can be administered.

CHAPTER 4. INSTRUMENTATION

136

Both the relatively low sensitivity of hybrid systems and the count rate

limitations result in the need for relatively long acquisition times, particularly

for whole body imaging studies. A further constraint is the period required to

measure attenuation in these studies. This makes the total time required for

whole body acquisition a critical factor in determining the utility of a system. In

addition, since iterative reconstruction is commonly used instead of filtered

back-projection, processing can be relatively slow. The total time of

examination including processing should be taken into consideration.

(d) Flexibility

Despite some constraints on performance, hybrid systems are consid-

erably more flexible than dedicated systems. This can be a major consideration

in situations where patient numbers or radionuclide supply may be limited.

Hybrid systems have considerable appeal for certain centres. New develop

ments in detector technology are likely to result in a wider range of hybrid

systems.

It should be noted that the technology used in dual photon imaging is

changing rapidly. As it develops, additional factors might need to be taken into

consideration.

4.4.7. Specification

A new NEMA publication (Performance Measurements of Positron

Emission Tomographs (see the bibliography to this section)) has been finalized

on the specification of dual photon imaging systems and is applicable to both

conventional PET systems and gamma camera based coincidence detection

systems. The emphasis of this document is on instruments designed for whole

body applications, although additional tests are included that provide

comparative information related to other types of application. The major

advance in this document is that no distinction is made between conventional

and gamma camera based systems. A more direct comparison between the

specifications should therefore be possible in the future.

The parameters specifically defined in the new document include those

listed below:

4.4. DUAL PHOTON IMAGING

137

(a) Spatial resolution

The report includes the radial and tangential FWHMs at the centre and

10 cm off-axis as well as the axial resolutions at the same positions.

(b) Scatter fraction, count losses and random events

This includes count rate plots for true, random, scatter and total events as

well as noise equivalent count rate. Peak counts are specified for each type of

event as well as scatter fraction.

System sensitivity is specified, with extrapolation to a zero attenuation

situation. The axial sensitivity profile is also presented.

(d) Accuracy — corrections for count losses and random events

The deviance of the corrected count rate from the expected true count

rate is specified, as are image quality and accuracy after attenuation and scatter

corrections. The contrast is calculated for spheres in a whole body phantom.

The additional tests suggested for applications other than whole body

studies are:

Scatter fraction

Count loss and random event measurements (dead time and true event

rates) should be made.

4.4.8. Acceptance testing

As in the case of single photon imaging, it is important that all aspects of

system performance are tested immediately after installation, and the ability of

the system to meet the functionality standards specified in the purchasing

document must be confirmed. These include:

—Sensitivities in 2-D mode (with interplane septa) and 3-D mode (with no

interplane septa);

—Energy resolution;

—Width of the coincidence time window;

—Spatial resolutions in the transaxial and axial directions;

CHAPTER 4. INSTRUMENTATION

138

—Variation in the detection sensitivity in the detection volume,

—True random ratio;

—Scatter fraction;

—Accuracy of scatter and attenuation correction.

4.4.9. Quality control

As in the case of SPECT there is a need for quality control procedures on

a daily basis, at regular intervals and on an as needed basis.

The daily quality control procedures include:

—Checking detector performance with a standard source;

—Updating detector normalization;

—Monitoring and recording any shift in parameters and environment.

The regular quality control procedures include:

—Setting up and recalibrating the detector;

—Checking the working parameter setting of the device;

—Making a phantom study of transmission and emission.

The less frequent quality control tests include:

—After power shutdown: checking detector set-up and normalization;

—After servicing: checking detector set-up, performance and normali-

zation;

—After change of source: checking normalization and making a phantom

study;

—When necessary: changing the transmission sources.

4.5. OTHER INSTRUMENTATION

4.5.1. Radiation protection and measurement equipment

Any nuclear medicine facility involves the use of radiation in many

different ways, including:

—Handling, storage and disposal of small to large activities of radioactive

material, potentially in gaseous, liquid and solid forms;

—Storage and handling of sealed radiation sources;

4.5. OTHER INSTRUMENTATION

139

—Monitoring of radiation levels in the work environment;

—Dealing with accidents involving any source used;

—Therapy with radioactive materials.

Even small facilities involve most, if not all, of the above. As a result,

different types of radiation measuring equipment are required as follows:

—Passive personnel dosimeters;

—Active (direct reading) personnel dosimeters;

—Contamination monitoring instruments (photons and beta radiation at

least);

—Radiation field monitoring instruments (photons).

4.5.1.1. Types of radiation detectors

The various types of radiation detector are described briefly below, in

particular their advantages, disadvantages and uses, all of which must be

understood by the user.

(a) Gaseous detectors — Geiger counters

Common, rugged, relatively cheap and reasonably accurate, the Geiger

counter is probably the most versatile detector available for nuclear medicine.

It operates by measuring individual radiation events, which can also be

smoothed out into a continuous signal of radiation exposure rate. Geiger

counters can be calibrated to read in units of absorbed dose or equivalent dose,

with, however, limited accuracy.

The detector itself is usually in the form of a cylinder of varying size, from

2 cm long by less than 1 cm diameter, to around 10 cm long by 3 cm diameter.

The detector may have a thin entry window for more efficient detection of low

energy photons and particles. Geiger counters are characterized as side

window, end window or pancake types. The first two usually have a shield to

filter out particles so that only photons are measured. Removing the shield also

allows particles to be detected, for example in contamination measurements.

End window type detectors can be made very thin to allow beta and even

alpha particles with energies greater than about 50 keV to be detected, whereas

side window types (with a larger surface area) are thicker and will only allow

photons and more energetic beta particles to pass. Pancake type detectors also

have a thin window but a larger area, and are designed for contamination

measurements.