Logo of the Physikalisch-Technische Bundesanstalt

Calibration of ionisation chambers for dose measurement in computer tomography


Approx. 6% of all X-ray examinations in 2003 in Germany were computer tomography examinations (CT) which, however, are responsible for more than 50% of the effective collective dose caused by X-ray diagnostics altogether in the population. These figures emphasise the high significance of dose measurements in CT. The most widely used measurand is thereby the so-called "CT dose index" (CTDI) which is measured by means of CT ionisation chambers calibrated in the measurand "air kerma-length product" in the unit mGy•cm. Cylindrical CT ionisation chambers have a diameter of approx. 10 mm and lengths of 10 to 15 cm. By means of calculated conversion factors, it is possible to determine the effective patient dose on the basis of the air kerma-length product. The calibration of such chambers is carried out in two steps: First, the mean air-kerma rate is measured at a reference level vertically from the axis of an X-ray beam by means of a standard chamber. The pencil-shaped CT chamber is then set up with its axis in such a way that the beam axis meets the chamber axis in the centre of the chamber. The average chamber current I is measured by irradiating the chamber at the known air-kerma rate over a defined length L (see Figure 1). The calibration factor is then calculated on the basis of the quotient of the product of the air-kerma rate and the length of the irradiated chamber, and the ionisation current of the chamber in the unit Gy/s•cm/A or Gy•cm/C. At PTB, the beam is blocked by means of lead plates of 4 mm thickness and a slit width of 50 mm. 

Figure 1 : Set-up for the measurement of the CT ionisation chamber current as a function of the irradiated chamber length L using a lead screen with variable aperture.

Bochud et al. [1] carried out measurements to optimise the calibration procedures for CT chambers. For this purpose, they used a 120 kV X-ray radiation which is representative of CT imaging and was filtered by means of 2.5 mm aluminium. In order to determine the influence of scattered radiation at the borders of the lead screen on the calibration factor, the CT chamber current I was measured as a function of the irradiated length L of the CT chamber. The irradiated length L was varied by modifying the slit aperture of the lead screen. A linear fit at the points of measurement I(L) of the currents I as a function of the irradiation lengths L demonstrated that the current I0 which is interpolated with the irradiation length L=0 does not yield 0. Furthermore, I0 was more than one order of magnitude higher than the measured current of the chamber at a same X-ray radiation but by shielding the chamber completely using lead. The following conclusion was drawn: I0 is the ionisation current which is caused at the borders of the screen by scattered radiation. In order to correct this undesirable influence, I0 was subtracted from the measured values I(L) so as to determine the calibration factor on the basis of the corrected values. The relative share of I0 on I(L) naturally decreases when the irradiation length L is increased; however, it was still of 26% at L=10 mm and 4% at L=50 mm. As mentioned above, at PTB, CT ionisation chambers are calibrated at L=50 mm and with a relative expanded measurement uncertainty of 1.5%. The influence of the scattered radiation at the level of the lead screen on the calibration factor in this measurement geometry has been determined as being smaller than 0.5% by means of Monte-Carlo simulations. Time-consuming measurements of this influence were therefore spared on the occasion of routine calibrations. However, since the results of the Monte-Carlo simulations deviated significantly from the measurement results yielded by Bochud et al. [1], it was necessary that the Department carried out their own measurements in this respect.

In order to clarify the situation experimentally, the influence of the scattered radiation was measured at PTB according to the method used by Bochud et al. [1] under the same conditions, i.e. for a 120 kV X-ray radiation with 2.5 mm aluminium filtering (Figure 1 shows a scheme of the experimental set-up). Figure 2 lists the results of the measurements of the CT chamber current standardised on a monitor chamber current as a function of the irradiated length L at the measurement location of the CT chamber. The value I0 which was determined on the basis of the linear fit was of 1.8% related to the current at L=10 mm and of 0.3% at L=50 mm and thus, it lies within the uncertainties that had been estimated by PTB with regard to this effect. The necessary correction of the ionisation current caused by scattered radiation in the CT chamber is thus smaller by one order of magnitude than that measured by Bochud et al. It is presently being investigated how it could happen that Bochud et al. overestimated this effect in such a way. Such investigations are of great importance in practice, e.g. in "Dosimetry in Diagnostic Radiology: An International Code of Practice" [2] or for the measurement procedures for checking the requirements on such chambers [3].

Figure 2 : The current ICT of the ionisation chamber related to the current IMK of the monitor chamber as a function of the irradiated chamber length L. The measured values for L = 10 mm to 90 mm are shown in intervals of 10 mm () and a linear fit on the measured values is represented by ().


  1. Bochud, F.O; Grecescu, M. and Valley, J-F.:
    Calibration of ionization chambers in air kerma length.
    Phys.Med.Biol. 46, 2477-2487 (2001).
  2. Dosimetry in Diagnostic Radiology:
    An International Code of Practice
    (IAEA Technical Reports Series No. 457, STI/DOC/010/457), 2007, in press
  3. IEC 61674 1997 International Standard:
    Medical Electrical Equipment-Dosimeters with Ionisation Chambers and/or Semi-Conductor Detectors as used in X-ray Diagnosis Imaging
    (Geneva: International Electrotechnical Commission)