chapter 8 Electronic Instrumentation for Radiation Detection Systems

Most of the radiation detectors used in nuclear medicine are operated in a “pulse mode”; that is, they generate pulses of electrical charge or current that are counted to determine the number of radiation events detected. In addition, by analyzing the amplitude of pulses from the detector, it is possible with energy-sensitive detectors, such as scintillation and semiconductor detectors and proportional counters, to determine the energy of each radiation event detected. Selection of a narrow energy range for counting permits discrimination against events other than those of the energy of interest, such as scattered radiation and background radiation or the multiple emissions from a mixture of radionuclides.

Figure 8-1 shows in schematic form the basic electronic components of a nuclear radiation-counting instrument. These components are present in systems ranging from the most simple sample counters to complex imaging instruments. The purpose of this chapter is to describe the basic principles of these components. The electronics for specific systems also are described in Chapters 10, 12 to 14, and 18. Basic principles of electricity and electronics are reviewed in the sources cited in the Bibliography at the end of this chapter.

FIGURE 8-1 Schematic representation of the electronic components for a nuclear radiation counting system.

a Preamplifiers

Table 8-1 summarizes the pulse output characteristics of detectors used in nuclear medicine. Most of them produce pulse signals of relatively small amplitude. In addition, most of the detectors listed have relatively high output impedance, that is, a high internal resistance to the flow of electrical current. In handling electronic signals, it is important that the impedance levels of successive components be matched to one another, or electronic interferences that distort the pulse signals may develop and system performance will be degraded.

TABLE 8-1 TYPICAL SIGNAL OUTPUT AND PULSE DURATION OF VARIOUS RADIATION DETECTORS

Detector	Signal (V)	Pulse Duration (µsec)
Sodium iodide scintillator with photomultiplier tube	10⁻¹-1	0.23*
Lutetium oxyorthosilicate scintillator with photomultiplier tube	10⁻¹-1	0.04*
Liquid scintillator with photomultiplier tube	10⁻²-10⁻¹	10⁻²*
Lutetium oxyorthosilicate scintillator with avalanche photodiode	10⁻⁵-10⁻⁴	0.04*
Direct semiconductor detector	10⁻⁴-10⁻³	10⁻¹-1
Gas proportional counter	10⁻³-10⁻²	10⁻¹-1
Geiger-Müller counter	1-10	50-300

* Mean decay time.

The purposes of a preamplifier (or preamp) are threefold: (1) to amplify, if necessary, the relatively small signals produced by the radiation detector, (2) to match impedance levels between the detector and subsequent components in the system, and (3) to shape the signal pulse for optimal signal processing by the subsequent components.

There are two main types of preamplifier configurations used with radiation detectors: the voltage-sensitive preamplifier and the charge-sensitive preamplifier. Figure 8-2 shows a simplified diagram of these two configurations. The symbol represents the signal (pulse)-amplifying component. The resistor (R) and capacitor (C) provide pulse shaping. The signal from the detector is typically a sharply rising pulse of electrical current of relatively short duration (1 µsec, except for Geiger-Müller (GM) counters; see Table 8-1). The voltage-sensitive preamp amplifies any voltage that appears at its input. Because radiation detectors are charge-producing devices, this input voltage, V_i, is given by the ratio of the charge, Q, and the intrinsic capacitance of the detector and other components in the input circuit, C_i:

FIGURE 8-2 A, Simplified circuit diagram of a voltage-sensitive preamplifier. The output voltage is determined by the amount of charge from the radiation detector, the input capacitance C_i, and the resistances R₁ and R₂. B, Simplified circuit diagram of a charge-sensitive preamplifier. The output voltage is determined by the charge from the radiation detector and the value of the feedback capacitor C_f. The symbol represents a voltage or current amplifying element. C, Input and output pulse signals for a charge-sensitive preamplifier. τ = (R _f × C_f) is the time constant of the pulse-shaping circuit.

(8-1)

With energy-sensitive detectors, the amount of charge, Q, and thus the amplitude of the voltage V_i are proportional to the energy of the radiation event detected. The output voltage, V_o, in this configuration is approximately

(8-2)

in which R₁ and R₂ are as shown in Figure 8-2. The minus sign indicates that the polarity of the pulse has been changed.

In semiconductor detectors, the input capacitance of the detector is sensitive to operating conditions, particularly temperature. Therefore the proportionality between charge and the voltage seen at the preamp input may not be stable. The charge-sensitive preamplifier overcomes this undesirable feature by using a feedback capacitor of capacitance C_f to integrate the charge from the radiation detector. The resulting output voltage, given by

(8-3)

is seen to be independent of the input capacitance, C_i.

The electrical charge leaks off the feedback capacitor through the resistor of resistance R_f, causing the voltage on the capacitor and at the outputs of the amplifier element to decrease exponentially with time t according to

(8-4)

The product R_f × C_f is called the time constant τ of the pulse-shaping circuit. The voltage decreases exponentially, dropping by 63% of its initial value during one time constant interval (see Fig. 8-2C). When R _f is given in ohms and C_f in farads, the time constant is given in seconds. Typical preamplifier time constants for nuclear medicine detectors (excepting those applications that require fast timing signals) are 20 to 200 µsec.

The amount of amplification provided by the amplifier element of the preamplifier varies with the type of detector. With scintillation detectors that use photomultiplier (PM) tubes, the PM tube already provides a considerable degree of amplification (10⁶-10⁷); thus relatively little additional amplification may be needed. Typically, a preamplifier gain factor (ratio of output to input amplitudes) of 5-20 is used for these detectors; however, some NaI(Tl):PM tube systems employ no preamplifier gain (gain factor of 1).

Detectors producing smaller signals, such as semiconductor detectors, may require a relatively high level of preamplifier gain, perhaps in the range of 10³-10⁴. It is not a trivial problem to design an amplifier that provides this amount of gain without introducing “noise signals” and temperature-related gain instabilities. Most of the modern high-gain preamplifiers employ field-effect transistors, which provide the desired low-noise and temperature-stability characteristics.

For energy-sensitive detectors, the preamplifier must operate in a linear fashion; that is, the amplitude of the signal out must be directly proportional to the amount of charge delivered to it by the detector. This preserves the relationship between pulse amplitude and energy of the radiation event detected, so that subsequent energy analysis may be applied to the pulse signals.

For the best results, the preamplifier component should be located as close as physically possible to the detector component. This maximizes the electronic signal-to-noise ratio (SNR) by amplifying the signal before additional noise or signal distortion can occur in the long cable runs that frequently separate the detector from the rest of the signal-processing components. This is particularly critical for detectors with small output signals (e.g., semiconductor detectors or scintillation detectors used for detecting low-energy radiations). It also is important for applications in which energy resolution is critical (see Chapter 10, Section B.7). Frequently, detectors and preamplifiers are packaged and sold as single units.

b Amplifiers

1 Amplification and Pulse-Shaping Functions

The amplifier component of a nuclear counting instrument has two major functions: (1) to amplify the still relatively small pulses from the preamp to sufficient amplitude (volts) to drive auxiliary equipment (pulse-height analyzers, scalers, etc.), and (2) to reshape the slow decaying pulse from the preamp into a narrow one to avoid the problem of pulse pile-up at high counting rates and to improve the electronic SNR.

The gain factor on an amplifier may range from ×1 to ×1000. Usually it is adjustable, first by a coarse adjustment (i.e., ×2, ×4, ×8) and then by a fine gain adjustment providing gain factors between the coarse steps. The coarse gain adjustment permits amplification of pulses over a wide range of amplitudes from different detectors and preamplifiers to the maximum output capability of the amplifier. The fine gain adjustment permits precise calibration of the relationship between amplifier output pulse amplitude (volts) and radiation energy absorbed (keV or MeV). For example, a convenient ratio might be 10 V of pulse amplitude per 1 MeV of radiation energy absorbed in the detector.

Pulse shaping—i.e., pulse shortening—is an essential function of the amplifier. The output of the preamp is a sharply rising pulse that decays with a time constant of about 50 µsec, returning to baseline after approximately 500 µsec. Thus if a second pulse occurs within 500 µsec, it rides on the tail of the previous pulse, providing incorrect amplitude information (Fig. 8-3). The system could not operate at counting rates exceeding a few hundred events per second without introducing this type of amplitude distortion.

FIGURE 8-3 Sequence of pulse signals in a radiation counting system. Top, Relatively long preamplifier time constant results in overlapping of pulse signals. Bottom, Amplifier output pulses have been shortened but without significant loss of amplitude or timing information.

The pulse-shaping circuits of the amplifier must provide an output of cleanly separated pulses, even though the output pulses from the preamp overlap. It must do this without distorting the information in the preamplifier signal, which is, mainly, (1) pulse amplitude (proportional to the energy of radiation event for energy-sensitive detectors) and (2) rise time (time at which the radiation event was detected). An additional function of the pulse-shaping circuits is to discriminate against electronic noise signals, such as microphonic pickup and 50- to 120-Hz power line frequency.

The most common methods for amplifier pulse shaping are resistor-capacitor (RC), gaussian, and delay-line methods. The RC technique, commonly referred to as RC shaping, is described to illustrate the basic principles. More detailed circuit descriptions are found in the sources cited in the Bibliography at the end of this chapter.

2 Resistor-Capacitor Shaping

Basic RC pulse-shaping circuits are shown in Figure 8-4. When a sharply rising pulse of relatively long duration (e.g., preamplifier output pulse) is applied to a capacitor-resistor (CR), or differentiation circuit (see Fig. 8-4A), the output is a rapidly rising pulse that decays with a time constant τ_d determined by the RC product of the circuit components (Equation 8-4). The amplitude of the output pulse depends on the amplitude of the sharply rising portion of the input pulse and is insensitive to the “tail” of any preceding pulse. Note that a CR differentiation circuit also is used for pulse shaping in the preamplifier; however, the time constants used in the preamplifier circuits are much longer than those used in the amplifier. Figure 8-4A also illustrates how the CR circuit discriminates against low-frequency noise signals.

FIGURE 8-4 Basic resistor-capacitor pulse-shaping circuits. A, Differentiation provides a sharply rising output signal that decays with time constant τ_d and discriminates against low-frequency noise. B, Integration circuit provides an output pulse that rises with time constant τ_d and discriminates against high-frequency noise.

Figure 8-4B shows an RC, or integration circuit. (Note that differentiation and integration differ only by the interchanging of the resistor R and the capacitor C.) When a sharply rising pulse is applied to this circuit, the output is a pulse with a shape described by

(8-5)

where V_i is the amplitude of the input pulse and RC = τ_i is the integration time constant of the circuit. This circuit discriminates effectively against high-frequency noise, as illustrated in Figure 8-4B.

Figure 8-5A shows a pulse-shaping circuit combining differentiation and integration stages. When the time constants of the two circuits are equal (τ = τ_i = τ_d), the output is a pulse that rises to a maximum value in a time equal to 1.2τ and then decays to approximately zero in 7τ. The maximum amplitude of the output pulse is determined by the amplitude of the input pulse. For scintillation and semiconductor detectors, a time constant in the range τ ~ 0.25-5.0 µsec usually is chosen. Thus the output pulse is shortened considerably relative to the pulse from the preamplifier (50-500 µsec) and is suitable for high counting rate applications. Except for a very small negative overshoot at the end of the pulse, the output pulse from this circuit has only one polarity (positive in Fig. 8-5A) and is called a unipolar output.

FIGURE 8-5 Resistor-capacitor pulse-shaping circuits combining differentiation and integration stages. A, Differentiation followed by integration. B, Differentiation-integration-differentiation circuit.

Figure 8-5B illustrates another type of shaping, called double differential shaping. The output pulse from this circuit has both positive and negative components and therefore is a bipolar pulse. For equal time-constant values, the bipolar output pulse has a shorter rise time and positive portion and a longer total duration than the unipolar output pulse. Unipolar pulses are preferred for signal-to-noise characteristics and are used where energy resolution is important. Bipolar pulses are preferred for high counting rate applications.

Research-grade amplifiers generally are provided with adjustable pulse-shaping time constants. A longer time constant provides better pulse amplitude information and is preferred in applications requiring optimal energy resolution, for example, with semiconductor detectors (Chapter 10, Section C.1). A shorter time constant is preferred in applications requiring more precise event timing and higher counting rate capabilities, such as scintillation cameras (Chapters 13 and 14) and coincidence detection of positron annihilation photons (Chapter 18).

3 Baseline Shift and Pulse Pile-Up

Baseline shift and pulse pile-up are two practical problems that occur in all amplifiers at high counting rates. Baseline shift is caused by the negative component that occurs at the end of the amplifier output pulse. A second pulse occurring during this component will be slightly depressed in amplitude (Fig. 8-6A). Inaccurate pulse amplitude and an apparent shift (decrease) in energy of the detected radiation event are the result (see Fig. 10-9).

FIGURE 8-6 A, Schematic representation of baseline shift, caused by a pulse riding on the “tail” of a preceding pulse. B, Pulse pile-up effects for two pulses occurring very close together in time.

Special circuitry has been developed to minimize baseline shift. This is called pole zero cancellation, or baseline restoration. This type of circuitry is employed in modern scintillation cameras to provide a high counting rate capability, particularly for cardiac studies. These circuits are described in the sources cited in the Bibliography at the end of this chapter.

At high counting rates, amplifier pulses can occur so close together that they fall on top of each other. This is referred to as pulse pile-up (Fig. 8-6B). When this happens, two pulses sum together and produce a single pulse with an amplitude that is not representative of either. Pulse pile-up distorts energy information and also contributes to counting losses (dead time) of the detection system, because two pulses are counted as one (Chapter 11, Section C).

Both baseline shift and pulse pile-up can be decreased by decreasing the width of the amplifier pulse (i.e., the time constant of the amplifier); however, shortening of the time constant usually produces poorer SNR and energy resolution. It is generally true that all the factors that provide high count rate capabilities in amplifiers also degrade energy resolution (Chapter 10, Section B.7).

Only gold members can continue reading. Log In or Register to continue