Reference dosimetry performed with calibrated ionization chambers in carbon beams currently still shows a significantly higher measurement uncertainty than that in high–energy photon fields. This is mainly attributable to a high uncertainty of the correction factor for the radiation quality kQ, which must be theoretically determined because there are no experimental data available.
In a previous project, it had already been possible to successfully determine kQ factors by experiment in the input channel of a monoenergetic carbon field. In this project, a reduction of the measurement uncertainty – compared to the theoretical values – by a factor of three was achieved.
This project is now continued with the determination of kQ factors in the spread-out Bragg peak (SOBP) of a carbon beam via water calorimetry. For this purpose, the SOBP is realized by a 2D range modulator. The resulting dose distribution was optimized and characterized in detail.
Due to the inverse depth dose curve and the intensity–modulated raster scan technique, the radiation therapy with high–energy ions applied at the Heidelberg Ion Beam Therapy Center (HIT) makes it possible to deposit the dose with high precision in the target volume whilst the surrounding tissue is spared.
For this spatially precise therapy method, precise dosimetry is, however, required in order to determine and verify the applied dose. The measurand in dosimetry is the absorbed dose to water which, in hospitals, is usually determined with the aid of calibrated ionization chambers.
Its application in carbon beams leads, however, to a significantly higher measurement uncertainty than in high–energy photon beams. With a relative standard measurement uncertainty of the dose determination of 3 %, the measurement uncertainty is about three times higher than in the case of photons [1]. This high uncertainty is largely determined by the high uncertainty of the kQ factor. The kQ factor describes the different responses of the ionization chambers to the radiation quality Q used, compared with the reference radiation quality Q0 (usually Co-60). Since there is no primary standard, the kQ factor for carbon beams is currently based on theoretical calculations.
Within the scope of a previous project [2], kQ factors have already been experimentally determined for the input channel of a monoenergetic carbon beam for two Farmer-type chambers. The result of this experimental determination was a total uncertainty of 0.8 %. This project is now being continued with the determination of kQ factors for the spread-out Bragg peak (SOBP) of a carbon beam. A possible change of the kQ factors as a result of the higher LET in this region – compared to the input channel – can be examined in this way.
Using the portable PTB water calorimeter [3], the absorbed dose to water is initially determined by measuring the radiation–induced temperature increase. Since the induced heat distribution gradually dissolves over time, the calorimetric measurements are relatively time–critical. Actively scanning the entire 3D field, which corresponds to an irradiation time of approximately 8 minutes for a dose cube with an edge length of 6 cm, would therefore lead to high uncertainties in the correction factors that are to be determined from the dose distribution.
The SOBP is therefore realized passively, using a 2D range modulator (2DRM, Fig. 1) that was created using 3D printing [4]. Therefore, only lateral scanning of the field is necessary. The irradiation time required for a dose cube of 6 x 6 x 6 cm³ is thus reduced to 90 s. The 2DRM consists of many pyramid-shaped pins whose shape is designed in such a way that both an SOBP with a depth dose profile as constant as possible and a lateral dose distribution as homogeneous as possible are generated.
Figure 1: Drawing of the 2DRM used with a 3 mm x 3 mm pin base area.
In particular, the influence which different base areas of the individual pins have on the resulting dose distributions was studied and optimized with respect to field homogeneity. For this purpose, base areas of 2 mm x 2 mm, 3 mm x 3 mm and 4 mm x 4 mm were tested.
The depth dose curve of a monoenergetic carbon beam that was modulated with the 2DRM shows the distinct formation of an SOBP (Fig. 2). However, two additional peaks at the beginning and at the end of the SOBP are visible which are caused by artifacts during the printing process. It became apparent that these artifacts diminish for larger pin base areas and even disappear for a base area of 4 mm x 4 mm. At the same time, with an increasing pin base area, the plateau area between the peaks flattens.
Measurements of the lateral dose distribution as a function of the pin base area, however, showed that it is more homogeneous for finer pins, i.e. smaller pin base areas. As a good compromise between axial and lateral homogeneity, a modulator with a 3 mm x 3 mm pin base area is used for the following measurements.
Figure 2: Depth dose curves of a 278 MeV/u carbon beam in water modulated with the 2DRM.
To be able to make a statement about the homogeneity and reproducibility, which form the basis for determining various correction factors, the dose distribution was thoroughly characterized.
By means of an ionization chamber array that can be remotely controlled at depth in a water phantom, three-dimensional measurements of the irradiation field were carried out. The setup used for this purpose – consisting of the water phantom, a waterproof casing for the array and a linear drive including control software – was developed and built at PTB. The time stability of the irradiation field was investigated by repeat measurements over a period of 10 weeks. The measured lateral dose distribution for one of these measurements is shown in Figure 3.
The measured data are each normalized to the mean value of the measured values within a sphere with a 2 cm radius around the center of the dose distribution and only the relative measured values are considered. For this area around the center, the relative standard deviation of the measured values from each other is 0.7–1.0 %. The relative deviation of the repeat measurements from each other is 0.3 %, which indicates good time stability.
Figure 3: Lateral dose distribution for different depths in the middle of the field over the x–axis (left) and y–axis (right), as measured with the ionization chamber array in the water phantom.
Under these optimized irradiation conditions, the first calorimetric measurements have already been performed with the portable PTB water calorimeter and two Farmer-type chambers. Further measurements are planned within the next months in order to be able to determine reliable values for the kQ factors and their overall uncertainties.
References:
(1) IAEA TRS-398, V.12 (2006)
(2) J.-M. Osinga-Blättermann et al, Phys. Med. Biol. 62 (2017) 2033–2054
(3) A. Krauss, Metrologia 43 (2006), 259-272
(4) Y. Simeonov et al, Phys. Med. Biol. 62 (2017), 7075-7096
Contact at PTB:
A. Krauss, Department 6.2, Working Group 6.23
K. Holm, Department 6.2, Working Group 6.23
