Physical foundations of non-invasive X-ray tube voltage measurement

The most important physical quantity, to be measured during development, servicing and QC of diagnostic radiology equipment is X-ray tube voltage. In the last decades several measuring instruments were issued with which X-ray tube voltage can be measured without breaking the high voltage circuit, by detection of X-rays only (non-invasive measuring devices).

From other parameters to be measured, non-invasive measurement of irradiation time is relatively easy. X-ray tube current can be measured “invasively” if X-ray generator has a transformer-middle output; however, in most recent X-ray generators generally it is not case. A “non-invasive“ X-ray tube current measurement is possible by a clamp, however, only if there is a high voltage cable, i.e. in case of monotank generators it is not possible. But both methods are outside of – in strict sense of the word – QA, basic principle of which is the exclusive non-invasivity of measurements. Therefore, in QA measurements, always displayed (nominal) value of X-ray tube current is considered as “true” value. Quantity (dose) of radiation is linearly proportional with the tube current time product while its quality (i.e. spectral distribution) – if tube voltage is constant – does not depend on it.

Quality of imaging influenced mostly by X-ray tube voltage as spectral distribution of the radiation – beside the absorbing media – is determined by it. At the same time, quantity of the radiation (dose) depends on it to an increased degree. Measuring of this high voltage – in the range of 40-150 kV, in mammography 20-50 kV – is complicated from technological point of view. In simpler, routine checks only peak value of the voltage, in more demanding cases also its waveform (as a function of time) is interesting.

X-ray tube voltage in the practice may be considered never as an absolutely smoothed direct voltage. That is the reason why for measuring X-ray equipment by conventional, “invasive” methods (i.e. by breaking the high voltage circuit) only ball spark-gap and high voltage ohmic dividers are used. With the former only the peak voltage can be measured while with the latter also waveform of the voltage. These methods are applied at manufacturing firms and type testing laboratories only.

In the recent decades non-invasive tube voltage measuring instruments (with which X-ray tube voltage can be measured without breaking the high voltage circuit, by detection of X-rays only) are quite widespread. These replace in many cases the complicated, time consuming and quite dangerous conventional methods. An even more important point of view is that in case of the so-called monotank X-ray generators (where high voltage transformer is placed into a housing common with the X-ray tube) measurement of X-ray tube voltage is possible only with non-invasive methods.

Spectrometric methods

 
As it is known for a long time, maximum photon energy and minimum wavelength of continuous spectrum of the X-ray beam are correlated with the peak X-ray tube voltage according to the so-called Duane-Hunt formula:

 
where e is the electron charge, h is the Planck constant, c is the velocity of light and U is the tube voltage. Measuring or U is given directly from (1). As values of all constants are known with very high accuracy, and there are no approximations, inaccuracy of determination of tube voltage depends only from inaccuracy of this measurement. With spectrometric methods an accuracy of some tenth kilovolt is reached. (More accurate measurement is possible with special, so-called calibrated voltage dividers only.) Measuring of X-ray spectra, however, demands a lot of space, very expensive measuring instrumentation, long time and high level professional expertise. So it is not appropriate for routine measurement, it can be recommended only for research purposes, type testing of monotank generators, and possibly calibration of other measuring instruments.

Radiographic methods

 
All radiographic methods use an absorbent stepwedge or wedge, placed onto a radiographic film. Taking and processing the radiogram, one has to find the step, density of which equals to that of the so-called reference strip. For every step there is a given range of kV for which this identity in density exists. This must – by the manufacturer – be calibrated with some other method. There is a difference between the individual devices in method of producing the constant density reference strip. These devices are called test-cassette.

In the two halves of the so-called Ardran-Crooks penetrameter cassette intensifying screens of different speed are put, the filter stepwedge is placed over the higher speed one, while the other produces the reference strip. With ageing of screens its accuracy decreases. – In the Wisconsin test cassette there is only one intensifying screen, the uniform density reference strip is produced by a light attenuating layer put into one half of the cassette. Nowadays it is manufactured by Gammex-RMI, a special version of it for mammography, too. – Principle of the so-called gyroscope test cassette is known for a long time, and it was produced in the 1980s in the former Soviet Union. In it the uniform density reference strip is produced by spinning a hole, cut in a lead layer.

Advantage of these methods is relative simplicity. Their disadvantage is the small accuracy (Wisconsin: ± 4 kV, gyroscope: ± 10 %) and the need for film processing. Therefore determining the value of kilovoltage in several points requires a lot of time and many radiographic films. These methods may be recommended to hospital technicians if no other, more up-to-date method is available. Nowadays, they are practically got out of use by the double-detector devices. With spreading of digital technology, even film processing possibility is not available in many places any more.

Double-detector devices

 
In 1932 L. Silberstein realized first that spectrum and attenuation curve (i.e. dose of transmitted beam as a function of absorber thickness) of an X-ray source is mathematically mutually related to each other, expressed with a Laplace transformation. Beam attenuation can be measured with the aid of a dosimeter and a series of absorbing material, from the results of which the spectrum can be calculated back and finally, from that, using (1), U is given. The calculation, however, is complicated. Simplification of this method could be possible by the recognition that even already two points of the curve are enough for assessment of tube voltage with acceptable accuracy.

Dosimetric measurements following each other as it was practiced earlier (e.g. R. H. Morgan 1944) is a significant origin of inaccuracy as actual tube voltage of the two exposures may be different. Therefore modern devices apply a detector arrangement with two detectors, covered by different (possibly changeable) filter thicknesses, and both detectors are to be irradiated simultaneously. Such devices were developed first in the USA in 1970s.

Accuracy of double-detector devices

 
Although the devices – according to conventional professional jargon – are called kVp (p: peak) meters, just difference of voltage (and current) from the peak value, i.e. influence of waveforms demands a well-established consideration. Non-microprocessor devices can measure only a time-averaged signal according to some preprogrammed calibration. Accuracy of determining peak kilovoltage by non-invasive methods is limited as only such differences of kilovoltage can be distinguished that cause higher differences in attenuation than the maximum ripple differences for the same peak kilovoltage. Calculations10,11 have shown that for kVp-measuring within 5-6 % accuracy two calibrations are sufficient: one of them for 1- and 2-pulse generators and the other for all other types (i.e. constant potential, medium frequency, 6- and 12-pulse). The accuracy may be increased by optimizing filter thicknesses.

Measuring instruments available on the world market

 
According to our best knowledge, there are only eight manufacturers throughout the world, producing non-invasive X-ray tube voltage measuring devices: in USA Fluke Biomedical (into this merged formerly Victoreen – with maintaining the trademark – and Keithley Radiation Instruments – without maintaining the trademark –), Gammex-RMI and Radcal, in Germany PTW-Freiburg, IBA Dosimetry (formerly Wellhöfer) and PehaMed, and in Sweden RTI Electronics and Unfors Instruments. There is a wide selection: the simplest device measures only kVp with one exposure (its controls are only an in/out knob and a “check” knob), while the most complicated microprocessor and with computer-connectable devices are able to measure many other parameters (invasive and non-invasive tube current and current time product, irradiation time, waveforms, dose and/or dose rate, relative dose/mAs linearity, number of pulses, dose/pulse, pulse rate, the same in fluoroscopic mode etc.). Devices of “middle category” measure – in addition to kV – irradiation time and a radiation (dose) waveform (in some cases also real-time kV-waveform, calculated with analogue method) is also given. Some of them are specialized to mammography or dental equipment. Manufacturers emphasize in some cases that device are applicable also for measuring high frequency generators, and the proper waveform type calibration is chosen automatically. Majority of most recent equipment also determines the filtration of the X-ray beam and measured kV-value is corrected with it. In case of recommended conditions most devices reach an accuracy of about 1,5-2 per cent.

For information, Figure 1. shows the simplified block scheme of the first double-detector microprocessor kV-measuring instrument, NERO (product of Victoreen).

Figure 1: Block scheme of NERO (Röntgensugárnyaláb-X-ray beam, Ionizációs karma – ionization chamber, réz szűrő-cooper filter, detector-detector, erősítők-amplifier, küszöb detector-threshold detector, mikroprocesszor-mikroprocessor)

 
In case of purchasing a measuring instrument, an important point of view for choosing – in addition to considering the area (radiography, fluoroscopy, mammography, dental, CT) – that the device should be suitable for quick and simple repair and regular recalibration, moreover, possibly in Europe.