First published online November 12, 2007
British Journal of Radiology (2008) 81, 46-50
© 2008 British Institute of Radiology
doi: 10.1259/bjr/71353258
Effects of phantom volume and shape on the accuracy of MRI BANG gel dosimetry using BANG3TM
N D MacDougall, MSc, MIPEM1,2,
M E Miquel, PhD1,2 and
S F Keevil, PhD, ARCP, FIPEM1,3
1 Division of Imaging Sciences, King's College London, Guy's Campus, London SE1 9RT, 2 Clinical Physics, St Bartholomew's Hospital, Barts and the London NHS Trust, London EC1A 7BE, 3 Department of Medical Physics, Guy's Hospital, Guy's and St Thomas' NHS Foundation Trust, London SE1 9RT, UK
Correspondence: Mr Niall D MacDougall, Head of Clinical Dosimetry, Radiotherapy Physics, Bart's and The London NHS Trust, 25 Bart's Close, West Smithfield, St Bartholomews Hospital, London EC1A 7BE, UK. E-mail: niall.macdougall{at}bartsandthelondon.nhs.uk
 |
Abstract
|
|---|
Complex radiotherapy techniques call for three-dimensional dosimetric methods with high spatial resolution. Radiation-sensitive polymer gel systems (e.g. commercially available BANGTM gel), read using MRI T2 mapping, offer a promising solution. A series of calibration test tubes is traditionally used to calculate the dose delivered to a larger, differently shaped volume of gel. In this work, we investigated the implicit assumption that the sensitivity of the gel is independent of shape and size. Phantoms of different shapes and volumes, and 20 glass test-tubes, were filled with BANG3TM gel. T2 mapping of gels was performed pre- and post-irradiation using a 32 echo Carr-Purcell-Meiboom-Gill sequence and single exponential fitting. Gel irradiation was performed with a 6 MV Varian 6EX linear accelerator. The T2 values of both non-irradiated and irradiated gels varied with container volume. For containers of the same shape receiving the same radiation dose, larger volumes exhibited a lower T2 value than did smaller volumes. Containers of the same volume but different shape also showed a smaller variation in response to radiation. The greatest difference in T2 values at the same dose was seen between test-tubes and larger volumes. This would imply that if test-tubes alone are used to calibrate larger volumes, then up to a 35% error could be introduced into radiotherapy plan verification. This can be reduced to <10% error if the gel volume is normalized with an external measurement device. Consequently, the traditional test-tube calibration method would be unacceptable for clinical plan verification.
|
Introduction
|
|---|
Newer and technically complex radiotherapy treatments, such as intensity-modulated radiotherapy and three-dimensional conformal radiotherapy, create the potential for higher rates of cancer cure [1]. However, the greater complexity increases the risk of treatment error and raises the need for more elaborate verification methods.
Ideally, radiation dose verification should be carried out in three dimensions, creating a demand for a new type of radiotherapy dosemeter. Polymer gels used in conjunction with a three-dimensional (3D) imaging modality, such as MRI, are a serious contender in the challenge to find a 3D tissue-equivalent radiation-measuring system [2]. Radiation-sensitive polymer gel systems (e.g. polyacrylamide gels [3]) and normoxic gels [4] are in increasingly widespread use for the verification of proposed radiotherapy treatment plans, particularly in complex dosimetric situations [5].
Most commonly, a volume of gel is irradiated in a dosimetry phantom and the transverse relaxation times (T2) of protons in the gel are mapped using MRI [3, 6]. Determination of radiation dose from this T2 map requires prior calibration, which is usually achieved by obtaining a reference curve from smaller gel volumes irradiated to a series of known doses.
When polymer gel dosimetry was first developed, it was commonly assumed that the response of the gel to radiation was linear [7], but it has subsequently been shown that this is not the case [8, 9]. Consequently, the shape of the radiation response curve must be characterized over the relevant dose range before meaningful dose data can be extracted from MRI results [10].
A common way to characterize this calibration curve is to irradiate a series of small volumes (usually test-tubes) of gel to different known doses and then to plot the dose against the R2 (1/T2) value of the samples. Other calibration methods such as percentage depth dose curves [11] can be utilized, but use of calibration samples that are typically of a different shape and size from the dosimetry phantom is common to all techniques. It is generally assumed that the calibration gels and the dosimetry phantom have the same radiation response (subject to inter-sample reproducibility).
Previous work has demonstrated the variation in radiation response among gel samples of identical size and shape [9]. Variation in response between test-tubes and a larger volume has been reported [8, 12], with differences of up to 22% noted. This work aims to quantify the variation in response due to both the size and the shape of gel volumes by measuring the radiation response of test-tubes and of larger volumes of gel in a variety of shapes and sizes.
|
Methods
|
|---|
Gel containers
20 test-tubes of spectrometer-quality glass (external radius 1 cm, volume 7 ml; Maram Instruments, Oxford, UK) were used. Other containers were manufactured using an acrylo-based resin (BAREX®, BP chemicals, London, UK). This acrylonitrile-methyl acrylate copolymer plastic is shatter-resistant and nearly impervious to oxygen, thus limiting any potential oxygen contamination. Each container was vacuum-formed in two halves, which were joined together using cyanoacrylic adhesive. Shapes manufactured were a set of cubes (volumes of 0.5 l, 1 l and 1.5 l), a sphere (1 l) and a rectangular parallelepiped (dimensions 1 l).
Gels
The gel used in this experiment was BANG3TM gel, the main ingredients of which are gelatine and methacrylic acid [13]. The gel was purchased as a single batch from MGS Research Inc (Madison, CT). Containers were filled by the company, and on arrival at our laboratory great care was taken to ensure that all samples had the same thermal and light exposure history. To ensure consistency, all gels were irradiated and then imaged in close succession. Except during scanning and irradiation, the gels were always kept in the same box to ensure as identical a physical history as possible.
T2 quantification
MRI was performed using a Philips Gyroscan Intera 1.5T MR system (Philips Medical Systems, Best, The Netherlands) with a bird-cage head coil. Transverse relaxation times were measured using a Carr-Purcell-Meiboom-Gill sequence (echo time (TE) = Nx15 ms for N = 1–32; voxel size = 5x1x1 mm; number of signal averages (NSA) = 2; repetition time (TR) = 1500 ms). Pixel-by-pixel single exponential fitting in Matlab (The MathWorks Ltd, Cambridge, UK) was used to generate T2 maps with every odd echo removed to lessen the effect of imperfect 180° pulses on the T2 measurement. The gels were imaged on delivery, before irradiation (first measurement) and 2 days after a selection of the gel samples had been irradiated (second measurement). The test-tubes were scanned along with each of the larger containers to allow correction for inter-scan variations.
Gel irradiation
Irradiation was performed on a Varian 6EX Linear Accelerator (Varian Medical Systems, Palo Alto, CA). Gel dosimetry experiments required delivery of a homogeneous radiation dose to each of the samples. The test-tubes were inserted into a specially produced 1.2 cm thick Perspex (poly(methyl 2-methylpropenoate); ICI, UK) slab to ensure adequate lateral scatter conditions. This was then sandwiched between two 5 cm blocks of WT1 water equivalent plastic (WEP; St Bartholomew's Hospital, London, UK [14]) to give adequate build-up and backscatter. The larger containers provided adequate scattering conditions to allow them to be irradiated without additional scattering material. Radiation field sizes were tailored to the size of each container (5x5 cm for the small cube and 8x8 cm for all others), and appropriate monitor units were set for the field size and depth required. The two larger cubes and the sphere were all irradiated with the same radiation field size.
The main factors affecting dose delivery accuracy in this setting were considered to be attenuation of the radiation field by the glass walls of the test-tubes, reduction in lateral scatter owing to the proximity of several test-tubes, and increased (compared with WEP) attenuation of radiation in Perspex. To investigate these effects, a calibrated ionization chamber (PTW-Freiburg Type 31010; 0.125 cm3) was placed in a test-tube in the experimental set-up and the dose delivered to the test tube volume was measured.
Before experiments began, the radiation output of the linear accelerator was characterized and verified. The output (cGy per monitor unit) was confirmed and the variation of dose in the plane perpendicular to the beam central axis (plane of gel irradiation) was determined. All radiation dose measurements were carried out with a National Physical Laboratory-calibrated 0.6 cm3 Farmer-type ionization chamber in conjunction with a Wellhofer electrometer, in accordance with the current UK high energy photon dosimetry code of practice [15].
Output was checked again before and after each experiment, and the calculated monitor units were adjusted to give the correct doses.
Each irradiation was carried out in two stages, with the direction of the radiation beam rotated through 180° between stages to ensure a homogeneous radiation dose distribution across the volume of gel. Irradiation doses for the different containers are summarized in Table 1
.
|
Results
|
|---|
Variation of T2 among gels between first and second measurements
T2 values of non-irradiated gel in the different containers were initially in the range 190–330 ms (first measurement). However, when the gels were re-imaged 2 days later (second measurement), the T2 values of gel samples that had not been irradiated (or had received negligible dose) were found to have increased to 320–370 ms (Figure 1). For the test-tubes and 0.5 l cubes, the second set of T2 values were obtained from non-irradiated samples. For the other containers, which had been irradiated in the interim, measurements were made in regions that had received very low doses (<3% of total dose). The same pattern of T2 increase is seen in both cases. For each set of measurements, T2 appears to increase as the volume of the container decreases, but is similar in containers of the same volume but different shape. The relative increase in T2 over time for each container was found to be proportional to volume.
View larger version (12K):
[in this window]
[in a new window]
|
Figure 1. Variation ofT2 with container size for volumes receiving zero dose or very low dose. For identical shape/volume combinations, at which more than one container was used (test-tubes: 20 points; small cubes: 3 points; medium cubes: 2 points; large cube: 2 points), the mean T2±standard deviation is displayed.
|
|
Radiation response of the gel
Figure 2 shows the variation in R2 value with dose for each container type detailed in Table 1
. The difference in R2 at "zero dose" reflects the variation in T2 seen in the second set of measurements. It is evident from this graph that, for containers of the same shape, larger volumes exhibit a higher R2 (lower T2) for a given dose than do smaller volumes. The R2 values of gel in differently shaped 1 l containers irradiated to the same dose are similar, but do vary slightly, with the rectangular parallelepiped showing the least response to radiation, followed by the sphere and then the cube (reacting most to the irradiation). The parallel opposed radiation delivery method used resulted in homogeneous regions (±1%) of radiation dose across the centre of the phantoms used. T2 values were taken from circular regions of interest (ROIs) with radius 1 cm in the phantoms, and from ROIs 0.6 times the internal radius of the test-tubes, to minimize the effect of Gibbs ringing artefacts [16].
View larger version (15K):
[in this window]
[in a new window]
|
Figure 2. Dose response of all containers. Error bars show limits ofR2 precision owing to the MRI scanning protocol.
|
|
Dependence of calibration accuracy on container volume
Test-tubes are frequently used to calibrate the dose distributions of larger volumes of gel. We can use the results presented here to investigate the errors arising from this procedure and how they vary with the volume of the dosimetry phantom. Figure 3 shows the percentage error that would arise if the test-tubes were used to calibrate the dose in each of the three sizes of cube. For example, if the R2 values for the large cube that received 1.5 Gy were converted to dose via the test-tube calibration curve, the measured dose would be 2.1 Gy — an overestimation of 40%. As can be seen in Figure 3, the accuracy of dose determination gets steadily worse with increasing dosimetry phantom volume. In addition, the percentage error is worse for lower doses.
View larger version (9K):
[in this window]
[in a new window]
|
Figure 3. Graph of percent dose error with volume when using test-tubes to convert volume MRI data to radiation dose.
|
|
|
Discussion and conclusions
|
|---|
Previous work has shown that the short-term precision of T2 measurements performed using the MR system and pulse sequence employed in this work is <0.5%, and the long-term accuracy of T2 measurement is <4%. Homogeneity of T2 measurements is better than 1%. There is no discernable radiofrequency heating- or temperature-induced changes in T2 as a result of the imaging pulse sequence employed [9]. The inherent uncertainty in T2 measurement arising from these potential sources of variation is too low to have any impact on the results presented in this paper. The radiation doses delivered to the phantoms were calculated taking into account the size and shape of the containers and the radiation delivery parameters used (e.g. field size, isocentre depth). The output of the linac was checked immediately before and after the experiment. As such, the radiation dose reproducibility was better than 1%.
The variation in T2 of non-irradiated gel appears to be well correlated with container volume. As the T2 of BANG gel is dependent on the amount of polymerization that has occurred, and because polymerization rate is temperature dependent, the slower cooling of larger volumes of gel may result in a larger proportion of polymerization and hence a lower T2. Also of interest is the rise in baseline T2 between the measurement taken before and after irradiation. If this was caused by polymerization, one would expect a fall in T2 values; one hypothesis for this is structural change in the gel owing to methacrylic acid reacting with the gelatin.
From the results shown in Figure 2, there is a strong suggestion that radiation response is related to gel volume. Gel container shape is another, but less important, variable. One of the factors underlying this behaviour may be the differences in the polymerization state of the non-irradiated gels discussed above. However, radiation-induced heating of the gel may also play a role [17] through the potential difference in relative cooling rates.
Previous work on a similar gel [5] did not show any differences in the T2 response of test-tubes with a range of cooling methods. However, the volume of gel in a container may be the main factor controlling the cooling rate, which would explain the differences between the radiation response of test-tubes versus larger gel volumes shown in this and previous work [8, 18].
The variation in T2 values observed between containers of identical volume but different shape are smaller and harder to explain. Differences in cooling rates may also be a factor but further work is required to investigate this fully.
It is evident from the results in Figures 2 and 3 that the test-tubes used here are not a suitable method of calibrating larger volumes of gel. One common approach to address this variation in dose response is to establish a point of known dose in the measured distribution using another dosimetry method, e.g. an ionization chamber or thermoluminescent dosemeter located in the dosimetry phantom. Armed with this information, one can normalize measurements made in the dosimetry phantom and the test-tubes at this point. Figure 4 shows the results of such an approach using 3 Gy as the normalization point (designated 100%). The relative accuracy of the gel dosimetry is improved close to the normalization point, with errors of ±5% in the 60–110% dose region. However, errors increase markedly at doses below 60% owing to the difference in the radiation responses in the calibration test-tubes and measurement phantoms.
View larger version (13K):
[in this window]
[in a new window]
|
Figure 4. Percent dose error when phantom dose distributions are normalized(3 Gy = 100%) at a known dose point.
|
|
These findings have significant implications for the use of gels in clinical radiotherapy dosimetry. The inherently different response of the dosimetry phantom and calibration gels means that measured doses may be in error over a large proportion of the dose range. The fact that low doses are overestimated will have implications for the verification of doses to organs at risk (OARs), which might in turn result in an unnecessary reduction in tumour dose and hence clinical efficacy. For example, in a head and neck treatment plan with an upper dose limit to the spinal cord of 48 Gy and a prescribed tumour dose of 68 Gy in 34 daily fractions, the gel dosimetric measurement would report the cord dose as 50.2 Gy. If the tumour dose was reduced in order to rectify this apparent overdose to the OAR, this would mean a 5% lower dose to the planning target volume, corresponding to a 10% reduction in the tumour control probability [19].
Received for publication February 2, 2007.
Revision received March 23, 2007.
Accepted for publication May 8, 2007.
|
References
|
|---|
- Zelefsky MJ, Fuks Z, Hunt M, Yamada Y, Marion C, Ling CC, et al. High-dose intensity modulated radiation therapy for prostate cancer: early toxicity and biochemical outcome in 772 patients. Int J Radiat Oncol Biol Phys 2002;53:1111–6.[CrossRef][Medline]
- DOSGEL. 3rd International Conference on Radiotherapy Gel Dosimetry. J Phys Conf Series 2004;3:228–31.[CrossRef]
- Maryanski MJ, Gore JC, Kennan RP, Schulz RJ. NMR relaxation enhancement in gels polymerized and cross-linked by ionizing radiation: a new approach to 3D dosimetry by MRI. Magn Reson Imaging 1993;11:253–8.[CrossRef][Medline]
- Fong PM, Keil DC, Does MD, Gore JC. Polymer gels for magnetic resonance imaging of radiation dose distributions at normal room atmosphere. Phys Med Biol 2001;46:3105–3.[CrossRef][Medline]
- Vergote K, De Deene Y, Duthoy W, De Gersem W, De Neve W, Achten E, et al. Validation and application of polymer gel dosimetry for the dose verification of an intensity-modulated arc therapy (IMAT) treatment. Phys Med Biol 2004;49:287–305.[CrossRef][Medline]
- Ibbott GS, Maryanski MJ, Eastman P, Holcomb SD, Zhang Y, Avison RG, et al. Three-dimensional visualization and measurement of conformal dose distributions using magnetic resonance imaging of BANG polymer gel dosimeters. Int J Radiat Oncol Biol Phys 1997;38:1097–103.[CrossRef][Medline]
- Maryanski MJ, Ibbott GS, Eastman P, Schulz RJ, Gore JC. Radiation therapy dosimetry using magnetic resonance imaging of polymer gels. Med Phys 1996;23:699–705.[CrossRef][Medline]
- Low DA, Dempsey JF, Venkatesan R, Mutic S, Markman J, Mark HE, et al. Evaluation of polymer gels and MRI as a 3-D dosimeter for intensity-modulated radiation therapy. Med Phys 1999;26:1542–51.[CrossRef][Medline]
- MacDougall ND, Miquel ME, Wilson DJ, Keevil SF, Smith MA. Evaluation of the dosimetric performance of BANG3 polymer gel. Phys Med Biol 2005;50:1717–26.[CrossRef][Medline]
- MacDougall ND, Pitchford WG, Smith MA. A systematic review of the precision and accuracy of dose measurements in photon radiotherapy using polymer and fricke MRI gel dosimetry. Phys Med Biol 2002;47:R107–21.[CrossRef][Medline]
- Oldham M, McJury M, Baustert IB, Webb S, Leach MO. Improving calibration accuracy in gel dosimetry. Phys Med Biol 1998;43:2709–20.[CrossRef][Medline]
- Farajollahi AR, Bonnet DE. A study of reproducibility and possible sources of error in gel dosimetry. Proc 2nd Int. Workshop on Radiation Therapy Gel Dosimetry; 2001.
- Maryanski MJ. Scavenging is no magic. Phys Med Biol 2002;47:L11–4.[CrossRef][Medline]
- White DR. The formulation of tissue substitute materials using basic interaction data. Phys Med Biol 1977;22:889–99.[CrossRef][Medline]
- Williams PC, Williams JR, Owen B, Lillicrap SC. Code of practice for high-energy photon therapy dosimetry based on the NPL absorbed dose calibration service. Phys Med Biol 1990;35:1355–60.[CrossRef]
- Stainsby JA, Wright GA. Partial volume effects on vascular T2 measurements. Mag Res Med 1998;40:494–9.[CrossRef]
- Salomons GJ, Park YS, McAuley KB, Schreiner LJ. Temperature increases associated with polymerization of irradiated PAG dosimeters. Phys Med Biol 2002;47:1435–48.[CrossRef][Medline]
- MacDougall ND, Miquel ME, Keevil SF, Smith MA. A dosimetric study of BANG gel calibration by T2 relaxometry. Proceedings of 12th ISMRM Annual meeting; 2004; Kyoto, Japan.
- IPEM. Balancing costs and benefits of checking in radiotherapy. Report No. 92.York: IPEM; 2006.
Top
Abstract
Introduction
