Basic Physics of Digital Radiography/The Source



The X-ray tube is, almost exclusively, the source of radiation beams used in Diagnostic Radiography. Fundamental features of this device and its application are considered in this chapter. Electrical generators which provide power to the tube and ancillary devices which can control the radiation beams are also described.

X-Ray Tube
X-rays can be generated by instruments such as the electron synchrotron and linear accelerator but in Diagnostic Radiography are nearly always produced by a small electron accelerator called an X-ray tube (XRT). We have described the basics of its operation in the previous chapter and will get into much more detail here.



An XRT in its simplest form consists of an anode and cathode mounted inside an evacuated glass tube - see Figure 2.1. The cathode generally consists of a small coil of wire, called a filament, mounted in a focusing cup. The anode target is generally made from tungsten separated from the filament by a small gap, although molybdenum and rhodium targets are used in Mammography. Electrons are boiled off the filament by applying an electric current so that it becomes white hot - the process is called Thermionic Emission. A high voltage (HV) of up to 100 kV or more is then applied and the electrons are attracted across the gap to collide with the anode at high energies, to produce X-rays. Bremsstrahlung and Characteristic Radiation processes are involved here, as previously described. The focusing cup can be used to form the electrons into a narrow beam and hence only strike a small spot on the anode target. X-rays are produced in all directions from this focal spot, but beam restriction devices (e.g. collimators; not shown in the figure) are used to allow only a primary beam to escape the source and irradiate the patient.

Most of the electron energy deposited in the anode is converted to heat with less than 1% actually producing X-rays. The focal spot therefore gets quite hot and it is here that the second important characteristic of tungsten comes into play. It has a high melting point (over 3,000oC) and therefore doesn’t melt under normal operating conditions.

Further points to note are firstly that the glass tube contains a vacuum so that electrons will not be absorbed and deflected by air molecules as they pass between cathode and anode. Secondly, the filament supply current is quite large (e.g. 5 A), but the current generated in the electron beam (the so-called X-ray tube current) is much lower (e.g. 1 mA in fluoroscopy to over 1,000 mA in angiography), depending on the efficiency of the thermionic emission. This current is one of the factors used to control X-ray exposure and its value is generally displayed on an ammeter (labelled mA in Figure 2.1). Thirdly, remember that ‘electron flow’ is the reverse of the direction of electric current (Michael Faraday got it wrong!), as indicated in the figure.

Modern diagnostic XRTs are more complex that the simple arrangement illustrated in Figure 2.1, mainly because fine focal spots are required to produce good spatial resolution in images (as discussed in the next chapter) and effective heat dissipation from the anode target is a major issue. Two design features have been incorporated into XRT design to address these conflicting requirements.



The first one is based on the Line Focus Principle and is illustrated in Figure 2.2. Here we see an exploded view of the anode surface being struck by an electron beam of height, ab. From below, the size of ab appears to be shortened to that of cd depending on the sine of the angle, θ, of the anode. The same consideration can be applied to the width of the electron beam, so that the heat it generates can as a result be dispersed over a broader region of the anode and a fine X-ray focal spot can still be obtained. The size of the electron beam focal spot is therefore reduced by angling the anode so that an apparently smaller X-ray source is obtained. This is called the effective size of the focal spot.

Anode angles are typically 6-15o in modern XRTs, with the small angled tubes being used for angiography, for instance, where fine detail imaging is required.



The second design feature to be incorporated is an anode rotated by an electric motor. This spins a disc of tungsten during the exposure so that the electron beam strikes an annular region of its surface rather than being concentrated into a small area, as in the stationary anode design we’ve considered up to now. The situation is illustrated in Figure 2.3 on the right where an end-on and side-on view of an anode disk is shown. Its diameter, R, can be up to 120 mm and it is generally spun at speeds of about 3,000 rpm during the exposure. A small effective focal spot can thus be obtained with the resultant heat being spread over of a broad area of anode material. The main advantage of the rotating anode XRT from a practical viewpoint is that short, intense exposures can be produced by a fine focal spot, without anode heating damaging the tube.

We are now in a position to consider the design of a rotating anode XRT - see Figures 2.4 and 2.5. The anode is generally made from a tungsten-rhenium alloy - the addition of a small amount of rhenium (e.g. 10%) having been found to reduce roughening of the anode surface during repeated use of the XRT. Many of these tubes can incorporate a second filament (not shown) so that both fine (e.g. 0.6 mm) and broad (e.g. 1.2 mm) focal spots can be obtained with the one XRT. A final point to note is that a thin exit window is used to reduce X-ray absorption and optimize the intensity of the X-ray beam used for the exposure of patients.

The process of taking an X-ray in diagnostic radiography therefore generally involves powering the motor until the optimal anode rotation speed is achieved, followed by the application of the HV to generate the X-ray beam. Simple, two stage switches are generally used for this purpose.



Two points should be noted before we finish. First, the angle of the anode presents a limit to the size of the field that can be covered. The smaller angles generate less coverage and the source-to-image distance (SID) therefore needs to be increased to compensate. The second point is that the angling of the anode causes differences in X-ray absorption to occur within the anode material itself so that (with reference to the Figure 2.6) X-rays emerging straight downwards, IA, will have an intensity lower than those exiting by a shorter route, IB. This causes a gradient in the X-ray intensities in the beam with lower intensities on the anode side - and is called the Heel Effect. The variation in intensity along the direction perpendicular to the anode-cathode axis is minimal in comparison. This can present challenges when using a small angled anode and can be compensated for, once again, by using a large SID.



The XRT is housed in a lead container (see Figure 2.7) so as to absorb the X-rays which are emitted by the anode in directions other than that of the patient beam. Some of these may actually stimulate secondary X-ray emissions from tube components other than the focal spot (called Off Focus Radiation) which the lead shielding is also designed to absorb. Radiation can nevertheless emerge from the tube housing that may lead to unnecessary irradiation of patients and staff. This leakage radiation should be kept to acceptable levels. The housing is also filled with oil which provides electrical insulation and assists in removing heat from the tube. It therefore typically contains a bellows or similar component which can provide for the consequent expansion of the oil. In addition, the housing generally contains an external marking which indicates the location of the focal spot.

Parameters which specify the range of application for an XRT include:


 * The Maximum Exposure Time for a single exposure given a particular kV and mA.
 * The Heat Capacity, i.e. the maximum amount of heat energy that can be stored in the anode.
 * Heating Curves which specify the length of time the anode can be heated at a range of kV and mA settings before the maximum heat storage capacity is reached.
 * Cooling Curves, which specify the time it takes an XRT to cool following use.

The maximum exposure time depends on factors such as the speed of the anode rotation, the size of the focal spot and the type of voltage waveform developed by the HV generator. Such information is generally provided in the form of Tube Loading Curves for each particular XRT. Curves are generally plots of the maximum exposure time for a given mA at a particular kV. An example is shown in Figure 2.7.5 where it can be seen that at 125 kV the maximum current that can be used for 0.1 s is approximately 330 mA whereas at 80 kV it is 500 mA.



The amount of energy, E, deposited in the anode is generally expressed as:

E = kV x mA x exposure time.

When the exposure time is measured in seconds, the energy deposited is expressed in joules. Anode heat capacities generally range from 200 kJ for mammography XRTs, through to 250 kJ for general radiography tubes, to 750 kJ for angiography tubes. Finally, the Power Rating of an XRT is defined as the maximum power in kilowatts that can be applied to the tube for 0.1s, i.e. the product of kV and mA divided by 1,000. Typical values are generally between 40 and 80 kW.

Recent developments in the design of X-ray tubes are reviewed in Behling (2016).

X-Ray Energy Spectrum
The form of the X-ray energy spectrum generated by an XRT is shown in Figure 2.8. It can be seen that it differs from spectra generated at the anode itself (i.e. the inverse linear dependence we saw in the previous chapter) because it is modified by the absorption characteristics of the glass wall, the oil, the housing exit port (e.g. bakelite) and any filters added at the XRT output. This energy filtration results in the preferential removal of low energy X-ray photons, which would not normally be transmitted through the patient to form an image and hence contribute an unnecessary radiation dose - and will be considered in more detail later in this chapter.



The spectrum for the 100 kV beam is seen in the figure to consist of a broad Bremsstrahlung spectrum with superimposed K-characteristic lines, as previously, but this time with low energy X-rays removed by the filtration. The beam is said to have been hardened by the filters, i.e. the average beam energy is increased. The 60 kV spectrum is seen in comparison to be of much lower intensity and to have insufficient energy to generate any K-Characteristic Radiation in the tungsten atoms.

The number of X-ray photons generated at the anode depends on the X-ray tube current, i.e. the mA. The effect of mA on the energy spectrum is to increase the intensity at all X-ray energies, and to otherwise leave the spectrum unchanged. This is illustrated in Figure 2.9 for the case of a 60 kV beam. Note that it is the product of tube current (mA) and the exposure time (s), i.e. the mAs, expresses the X-ray intensity generated in recording a radiograph.



The number of X-ray photons generated at the anode is also strongly dependent on the applied voltage (i.e. the kV), with an approximately square relationship. The efficiency of an XRT for X-ray production is therefore substantially greater at the higher kilovoltages. This is illustrated in Figure 2.9.5, where the number of photons produced at 60 kV is seen to be about 2.5 times the number produced at 40 kV, while the number produced at 100 kV is almost 7 times that amount - and the number at 140 kV almost 12 times. It is important to appreciate that both the mA and the kV affect the number of X-ray photons in a beam.



Note that since a broad X-ray energy spectrum is generated by the XRT, the emitted X-rays are referred to as polychromatic or polyenergetic, in contrast to monochromatic (or monoenergetic) emission, which refers to X-rays having a single energy.

Note also that, since most radiation detectors are far too sensitive for measuring X-ray energy spectra directly, computer-based simulators, e.g. SpekCalc, can be used as an acceptable alternative.

HV Generator
This is an electrical device for generating the high voltage (HV) necessary to power the XRT. Mains electricity is generally supplied to a hospital/clinic in the form of an alternating current (AC) in single-phase or three phases. A single-phase HV generator can take the power supplied by the AC mains, rectify it and transform it to tens of thousands of volts prior to application to the XRT. The rectification process is used, either in half-wave or full-wave form, to remove any negative going voltages which would otherwise excite the anode to act as an electron source (see the Figure 2.10).



Note from the figure that half-wave rectification causes a voltage pulse to be applied to the XRT every mains cycle with a duration equivalent to half a mains cycle (e.g. 10 ms in Europe and Australia). The voltage pulse is also seen to rise from zero, reach a maximum and head to zero again. It is apparent that the maximum energy of the X-rays produced will therefore follow a similar pattern. This maximum voltage is referred to as the peak kilo-voltage (kVp) as a result.

It is also seen in Figure 2.10 that full-wave rectification will provide two voltage pulses every mains period, but that the fluctuation will still remain. Note however that the length of the exposure (i.e. the exposure time) can ideally be halved to achieve an equivalent X-ray exposure as a half-wave rectified system.

It is important to appreciate that capacitive effects in the HV cables can have a smoothing effect on these voltage fluctuations, so that the situation is not quite as simple as that described here.

These large fluctuations in voltage can be reduced to ripples using more sophisticated generators. For example, generators powered by three-phase electricity can produce 6 and 12 pulses per mains cycle, with voltage ripples of 13.5% and 3.5% respectively. Ripple can be reduced even further, to 0% ideally, in the highly-specialized constant potential generator.

Inverter HV generators have more recently been developed. These consist of circuitry which invert the incoming AC to DC before chopping it into a medium frequency (e.g. 2 kHz) or high frequency (e.g. 200 kHz) and then transforming this to a kilo-voltage. Such generators can produce a voltage ripple of about 5%.

Since these modern generators produce a virtually constant high voltage, the X-ray spectrum remains constant during the exposure, unlike single-phase generators. The term kVp has therefore been replaced by the term kV.



Mobile radiography systems can be powered by such inverter circuits, or, more traditionally, by Capacitor Discharge generators. These older systems work by charging a capacitor to the desired kV and discharging it through the XRT. An exponentially (ideally) decreasing voltage is therefore applied to the XRT during these exposures.

The generator console usually provides protocols to be used for radiography different body parts - see Figure 2.11. The exposure control is generally located here and the kilo-voltage (kV), tube current-exposure time product (mAs) and exposure time (ms) for the exposure are typically displayed.

Beam Filtration
Filtration is the process of preferentially removing low energy X-ray photons from a polychromatic radiation beam. Its general purpose is radiation protection because these low energy photons are likely to be completely absorbed in the patient and not contribute to image formation. X-rays at all energies are actually attenuated by the filtration process with the lower energies being suppressed to a greater extent. The process therefore hardens the radiation beam, i.e. the X-ray spectrum is shaped so that its mean energy increases - see Figure 2.12.



Thin sheets of aluminium and copper have traditionally been used for such filtration. Rare earth filters have also been used. In the case of lanthanum, which has a K-edge at 39 keV, there is a marked increase in attenuation at this energy, and the resultant spectrum is shaped as illustrated in the figure.

These metal filters are generally referred to as Added Filtration. They’re to be contrasted with the Inherent Filtration of the XRT and housing, described earlier in this chapter, provided by the glass of the tube wall, the oil and the bakelite exit port. These components can generate a filtration equivalent to about 0.8 mm Al, typically.

The Total Filtration of an X-ray beam is therefore the sum of the inherent and added filtration. The application of added filters requires an increase in exposure factors to compensate for their X-ray absorption. In general, a total filtration equivalent to at least 2.5 mm Al is used for exposures greater than 70 kVp, with thinner thicknesses being used for lower energy exposures.

Light Beam Diaphragm
A light beam diaphragm (LBD) is generally attached at the output port of the XRT to allow the beam dimensions to be adjusted using collimators. Typically, these consist of two sets of horizontal lead plates that can be adjusted so that a rectangular radiation beam can be formed - see Figure 2.13.



The device also generally contains an off-axis light globe with a small mirror mounted along the X-ray beam axis to reflect the light so that it mimics the dimensions of the X-ray beam. The light beam is used to set-up the field to be exposed at the patient prior to X-ray exposure and generally has crosshairs superimposed.

A slot in the LBD can be noted in the photo. This is used for the placement for added filtration, as previously discussed.

Position controls are generally located at the LBD and allow the XRT to be moved vertically and horizontally as well as to be angled for oblique exposures. The controls for adjusting the collimators are also generally located at the LBD - see Figure 2.14.



The LBD is also an ideal location for a dose-area product (DAP) meter to be mounted to monitor radiation exposures to patients.

Automatic Exposure Control


Automatic exposure control (AEC), also called Photo-Timing, is widely used in Diagnostic Radiography to terminate X-ray exposures when a reference level is reached. This reference is generally pre-defined by the exposure required at the image receptor to produce adequate radiographs. Radiation detector(s) are placed at the image receptor to measure the exposure and feed a signal back to the HV generator when termination is required.

The detectors can be based on ionization chambers which can be mounted on the anterior side of the image receptor because of their high X-ray transparency. They can also be based on solid-state detectors, but in this case are mounted posteriorly because of their opacity to X-rays. In addition, a number of detectors can be used for different regions of the irradiated field. For instance, in chest radiography, three AEC detectors are typically used - one for the region of the mediastinum (indicated by the shaded rectangular region in Figure 2.15) and one for regions in each lung (indicated by the shaded circles). Exposure termination can be determined on this basis in PA chest radiography using the average exposure determined from the two lung detectors, and using the mediastinal detector for imaging the thoracic spine. Other detector configurations are also possible, depending on the clinical application.

The process of taking a radiograph when using an AEC therefore includes the following:


 * selecting the appropriate kV on the console;
 * powering the XRT motor;
 * applying the exposure until automatic termination of the exposure.

Note that releasing the exposure switch too early can terminate the exposure prematurely on some systems, although this is rarely an issue given the very short exposure times used in most clinical exposures.

The mAs used to obtain the AEC reference level is displayed on the console following the exposure - see Fig. 2.11 above. The circuitry when AEC is used with a High Frequency HV generator is illustrated in Figure 2.16. Here, the incoming mains voltage is rectified and smoothed to DC voltage, before being converted to a high frequency AC voltage - which is then converted to a high voltage (HV) by a transformer, rectified and smoothed back to DC for application to the X-ray tube. For AEC operation, the detector voltage generated by the radiation exposure is integrated to give a measure of the actual exposure, Dactual, which is then compared to a previously-stored reference exposure, Dref to enable termination of the exposure.



Finally, this type of generator design can be extended to one which can generate radiographs with shorter exposure times. This can be achieved at a fixed kilovoltage and a current (mA) which falls from a higher value than normally used during the exposure in a manner which exploits the instantaneous heating characteristics of the XRT anode. These heating characteristics can be pre-programmed into the exposure control system. This type of design is referred to as a Falling Load generator and can be used to reduce exposure times to substantially less than those achievable with the more conventional design.