Interlaboratory comparison of two homemade pencil type ionization chambers

.


INTRODUCTION
The use of Computed Tomography (CT) for diagnostic images has shown a growing use, due to technological advances of this equipment [1][2][3].This type of equipment has undergone many improvements [4] with the aim to reduce the doses received by the patients during the procedure, because this kind of imaging procedure uses higher radiation doses, in comparison to other procedures of conventional radiology.Therefore, it is important to perform radiation dosimetry using appropriate and reliable detectors.
The dosimetry of CT beams is usually made using a pencil type ionization chamber with a sensitive volume length of 10 cm, but there are also other sizes available in the market.This detector presents a uniform response to the incident radiation beam from all angles, which makes it appropriate once the X-ray tube executes rotation movement around the patient table during irradiation [5].
The metrological reliability of a dosimetric detector is maintained through its calibration and performance tests.For this reason, it is very important that the equipment to be calibrated is traced to a reference instrument.
Traceability refers to how the response of a given detector is related to a given reference system.Therefore, there must be a chain in which all comparisons ensure the results of the measurements obtained, leading to the highest standard reference system, which in this case is the absolute detector of each type of radiation from the primary standard laboratory [6].The research group of the LCI has developed over the last ten years some prototypes of ionization chambers, including a pencil type [7][8][9][10].These homemade chambers were made with low cost material, and they present some small differences in relation to the commercial models.The aim of this work was to compare the response of two homemade pencil type ionization chambers with sensitive volume lengths of 10 cm and 30 cm at the LCI/IPEN and at the NPL, and as a consequence their calibration.

Equipment
The following equipment were utilized: • A secondary standard pencil type ionization chamber, Radcal, model RC3CT, with sensitive volume length of 10 cm and volume of 3 cm³, called CIPEN, of LCI.This chamber was calibrated at the Physikalisch-Technische Bundesanstalt (PTB) laboratory (Certificate n. 5889, 14/07/2009).
• A primary standard free air chamber utilized in the energy range of the laboratory radiation beams from 40 kV to 300 kV, called CNPL, of NPL.
• Two homemade pencil type ionization chambers, with sensitive volume lengths of 10 cm (called C10) and 30 cm (called C30) manufactured at IPEN. Figure 1 shows the design of the ionization chambers and Table 1 shows their specifications.
p. 5  For all of the measurements, both ionization chambers were connected to an electrometer, model UNIDOS E, Physikalisch-Technische Werkstatten (PTW), Germany.
The measurements obtained were corrected to the standard values of environmental temperature and pressure.The uncertainties of type A and type B were determined, with the combined uncertainty of factor k = 2.
The X-ray equipment, Pantak / Seifert, model ISOVOLT 160 HS, operating between 5 kV and 160 kV, was utilized at the LCI.In the National Physical Laboratory (NPL) the Xray equipment, Comet, model MXR-321, operating up to 320 kV was used.The characteristics of the CT standard X radiation qualities are presented in Table 2 (LCI) and

Characterization tests
The tests performed in both laboratories were: stabilization time, saturation curve, polarity effect, ion collection efficiency and linearity of response.
For the stabilization time test of the ionization chambers, the measurement standard deviation must not exceed ± 2 % when comparing the response ionization current obtained in 15 min and 60 min [12].
The polarity effect and the ion collection efficiency were obtained through the saturation curve.For the polarity effect test, the ratio of the ionization currents obtained for positive and negative polarities must not exceed ± 1 % [12]; therefore, the results should be in the range 0.99 -1.01.Using Equation 1 [6], it was possible to calculate the ion collection efficiency of each ionization chamber, where the standard deviation must not exceed ± 5 % [12].
where: M1 and M2 are the values obtained with the voltages V1 = + 200 V and V2 = + 100 V, respectively.
The linearity of response of both pencil type ionization chambers must present a linear curve, so the correlation factor needs to be close of 1.00 (R 2 = 1.00).
The two pencil type ionization chambers were calibrated in relation to the standard chambers (CIPEN and CNPL) using the replacement method.Through the results obtained, it was possible to determine the energy dependence, where the standard deviation must not exceed ± 5 % [12].

RESULTS AND DISCUSSIONS
For all tests each of the pencil type ionization chambers was positioned in a specific setup, at the distance of 1 m from the focus of the X-rays equipment used for calibration of radiation detectors.At the distance of 5 cm in front of the chamber there is a collimator with a central window with dimensions of 2 cm x 5 cm.For the tests, 5 measurements during 30 s were taken in the three radiation qualities.
Initially are presented the results obtained for the stabilization time test.After irradiation, the current was measured 15 min and 60 min after switching on the dosimetric system.The difference between results obtained in 15 min and 60 min was calculated and compared with the limits established by IEC (IEC, 1997), and they can be seen in Table 4.
p. 8 As can be observed in Table 4, the results obtained for the stabilization time test are within the recommended limit (± 2 %) in the three radiation qualities in both laboratories [12].
To determine the optimal voltage for the operation of the ionizing chambers, saturation curves were obtained varying the applied voltage to both ionization chambers from -400 V to + 400 V, in steps of 50 V.The results are shown in Figures 2 and 3 for the C10 and C30 chambers in both laboratories.As observed in Figures 2 and 3, the saturation curves present the same behavior for both polarities applied to the chambers C10 and C30 in all three CT radiation qualities in both laboratories.Comparing the chamber results obtained in the two laboratories, the only difference is the current ionization values, which is higher at the LCI, due to the higher air kerma rate.
The polarity effect was estimated by the calculation of the ratio between the ionization current values obtained for the positive and negative polarities, for each applied voltage utilized in the saturation curves.Tables 5 and 6 show the results obtained for the polarity effect at LCI and NPL for both ionization chambers.The results obtained for the polarity effect are within the recommended limits (± 1 %) [12] for both ionization chambers at the two laboratories, in the three CT radiation qualities.
As in the case of the polarity effect obtained from the saturation curve, the ion collection efficiency used the values obtained for the + 200 V and + 100 V in each radiation quality for both ionization chambers at LCI and NPL.The results are shown in Table 7.As observed in Table 7, the ion collection efficiency obtained of the chambers C10 and C30 in all three CT radiation qualities in both laboratories are within the recommended limits (± 95 %).
The linearity of response was obtained varying the applied current (tube current) to both ionization chambers from 2 mA to 20 mA at LCI and from 1 mA to 15 mA at NPL.
The results are shown in Figures 4 and 5 for the C10 and C30 chambers in both laboratories.As can be observed in Figures 4 and 5, each curve obtained for the linearity of response is linear; the correlation coefficients (R 2 ) are shown in Table 8.All of the correlation coefficients obtained for the two pencil type ionization chambers in the three CT radiation qualities in both laboratories (LCI/IPEN and NPL) are very close to 1.000.Therefore, it is possible to confirm the linearity of the response.
The energy dependence was determined for the two ionization chambers in both laboratories.It was obtained by the maximum deviation of the calibration factor for each radiation quality in relation to the correction factor obtained for the reference radiation quality (RQT 9).Tables 9 and 10 show the results obtained at LCI and Tables 11 and 12, at NPL laboratory.As can be observed, the energy dependence obtained for the ionization chamber C10 at LCI/IPEN and at NPL is within the international recommended limit of ± 5 % (IEC, 1997).The results obtained for the pencil type ionization chamber C30 in both laboratories are not in agreement with the international recommendations IEC 61674 [12].Table 13 shows the percentual difference between the calibration coefficients of chambers C10 and C30 in both laboratories.
p. 14 As can be seen in Table 13 for the C30 chamber the maximum deviation was less than 1%.However, for chamber C10 the highest difference was 15.2% for RQR 8; this difference may be due to the chamber design of the reference systems used for the calibration and the setup of the chamber during the measurements.

CONCLUSIONS
Is this study two homemade pencil type ionization chambers with sensitive volume lengths of 10 cm (C10) and 30 cm (C30), manufactured with low cost material, were tested in two laboratories (LCI and NPL) in the CT radiation qualities.The results obtained for the characterization tests (stabilization time, saturation curve, polarity effect, ion colleting efficiency and linearity of response) are all within the recommended limits.
Both chambers were calibrated at LCI and NPL laboratories, and the energy dependence of their responses was obtained.In the case of the C10 chamber, the results were within the recommended limit, but for the C30 chamber it was not the case, probably because of its extended chamber length.
The behavior of the two pencil ionization chambers showed to be similar in both laboratories; the only difference is that the ionization current obtained for each test is higher at the LCI/IPEN than at the NPL, due to the established radiation quality characteristics in each laboratory.

The
Instrument Calibration Laboratory (LCI) from Instituto de Pesquisas Energéticas e Nucleares (IPEN) has standard systems traceable to the Brazilian National Ionizing Radiation Metrology Laboratory (LNMRI) that is a secondary standard laboratory, and to the primary standard laboratories of the Physikalisch-Technische Bundesanstalt (PTB, p. 4 Germany), the National Institute of Standards and Technology (NIST, USA) and to the National Physical Laboratory (NPL, UK).

Figure 1 .
Figure 1.Design of a pencil type ionization chamber showing the components.

Figure 2 . 9 Figure 3 .
Figure 2. Saturation curves for the pencil type ionization chamber C10 in the three CT radiation qualities at (a) LCI and (b) NPL.The maximum measurement uncertainty was 0.05 %.

Figure 4 .
Figure 4. Linearity of response for the pencil type ionization chamber C10 in the three CT radiation qualities at (a) LCI and (b) NPL.The maximum measurement uncertainty was 0.03 %.

Figure 5 .
Figure 5. Linearity of response for the pencil type ionization chamber C30 in the three CT radiation qualities at (a) LCI and (b) NPL.The maximum measurement uncertainty was 0.03 %.

Table 1 .
Characteristics and specifications of the pencil type ionization chambers C10 and C30.The material of the collector electrode is aluminum with a layer of graphite, the chamber body material is PMMA with a layer of graphite and the cape material is Teflon.

Table 2 .
Characteristics of the CT standard X radiation qualities at the LCI (IEC, 2005).

Table 3 .
Characteristics of the CT standard X radiation qualities at the NPL (IEC, 2005).

Table 4 .
Stabilization time test for the pencil type ionization chambers (C10 e C30) in the CT radiation qualities in both laboratories (LCI and NPL).∆: Difference between the ionization current values measured in 15 min and 60 min.

Table 5 .
Polarity effect for the pencil type ionization chambers C10 and C30 in the CT radiation qualities at LCI.

Table 6 .
Polarity effect for the pencil type ionization chambers C10 and C30 in the CT radiation qualities at NPL.

Table 7 .
Ion collection efficiency (%) for the ionization chambers (C10 and C30) in the CT radiation qualities at the LCI and NPL.

Table 8 .
Correlation coefficients (R 2 ) obtained for the linearity of the response for both pencil type chambers (C10 and C30) in the three CT radiation qualities at LCI and NPL.

Table 9 .
Calibration coefficient, correction factor and energy dependence of the CT radiation qualities for the pencil type ionization chamber C10 at LCI.

Table 10 .
Calibration coefficient, correction factor and energy dependence of the CT radiation qualities for the pencil type ionization chamber C30 at LCI.

Table 11 .
Calibration coefficient, correction factor and energy dependence of the CT radiation qualities for the pencil type ionization chamber C10 at NPL.

Table 12 .
Calibration coefficient, correction factor and energy dependence of the CT radiation qualities for the pencil type ionization chamber C30 at NPL.

Table 13 .
Percentual difference on calibration coefficients for C10 and C30 in both laboratories.