Jan Vandecasteele

Evaluation of radiochromic gel dosimetry and polymer gel dosimetry in a clinical dose verification

A quantitative comparison of two full three-dimensional (3D) gel dosimetry techniques was assessed in a clinical setting: radiochromic gel dosimetry with an in-house developed optical laser CT scanner and polymer gel dosimetry with magnetic resonance imaging (MRI). To benchmark both gel dosimeters, they were exposed to a 6 MV photon beam and the depth dose was compared against a diamond detector measurement that served as golden standard. Both gel dosimeters were found accurate within 4% accuracy. In the 3D dose matrix of the radiochromic gel, hotspot dose deviations up to 8% were observed which are attributed to the fabrication procedure. The polymer gel readout was shown to be sensitive to B0 field and B1 field non-uniformities as well as temperature variations during scanning. The performance of the two gel dosimeters was also evaluated for a brain tumour IMRT treatment. Both gel measured dose distributions were compared against treatment planning system predicted dose maps which were validated independently with ion chamber measurements and portal dosimetry. In the radiochromic gel measurement, two sources of deviations could be identified. Firstly, the dose in a cluster of voxels near the edge of the phantom deviated from the planned dose. Secondly, the presence of dose hotspots in the order of 10% related to inhomogeneities in the gel limit the clinical acceptance of this dosimetry technique. Based on the results of the micelle gel dosimeter prototype presented here, chemical optimization will be subject of future work. Polymer gel dosimetry is capable of measuring the absolute dose in the whole 3D volume within 5% accuracy. A temperature stabilization technique is incorporated to increase the accuracy during short measurements, however keeping the temperature stable during long measurement times in both calibration phantoms and the volumetric phantom is more challenging. The sensitivity of MRI readout to minimal temperature fluctuations is demonstrated which proves the need for adequate compensation strategies.

On the validity of 3D polymer gel dosimetry: III. MRI-related error sources

In MRI (PAGAT) polymer gel dosimetry, there exists some controversy on the validity of 3D dose verifications of clinical treatments. The relative contribution of important sources of uncertainty in MR scanning to the overall accuracy and precision of 3D MRI polymer gel dosimetry is quantified in this study. The performance in terms of signal-to-noise and imaging artefacts was evaluated on three different MR scanners (two 1.5 T and a 3 T scanner). These include: (1) B₀-field inhomogeneity, (2) B₁-field inhomogeneity, (3) dielectric effects (losses and standing waves) and (4) temperature inhomogeneity during scanning. B₀-field inhomogeneities that amount to maximum 5 ppm result in dose deviations of up to 4.3% and deformations of up to 5 pixels. Compensation methods are proposed. B₁-field inhomogeneities were found to induce R₂ variations in large anthropomorphic phantoms both at 1.5 and 3 T. At 1.5 T these effects are mainly caused by the coil geometry resulting in dose deviations of up to 25%. After the correction of the R₂ maps using a heuristic flip angle-R₂ relation, these dose deviations are reduced to 2.4%. At 3 T, the dielectric properties of the gel phantoms are shown to strongly influence B₁-field homogeneity, hence R₂ homogeneity, especially of large anthropomorphic phantoms. The low electrical conductivity of polymer gel dosimeters induces standing wave patterns resulting in dose deviations up to 50%. Increasing the conductivity of the gel by adding NaCl reduces the dose deviation to 25% after which the post-processing is successful in reducing the remaining inhomogeneities caused by the coil geometry to within 2.4%. The measurements are supported by computational modelling of the B₁-field. Finally, temperature fluctuations of 1 °C frequently encountered in clinical MRI scanners result in dose deviations up to 15%. It is illustrated that with adequate temperature stabilization, the dose uncertainty is reduced to within 2.58%.

On the validity of 3D polymer gel dosimetry: I. reproducibility study.

The intra- and inter-batch accuracy and precision of MRI (polyacrylamide gelatin gel fabricated at atmospheric conditions) polymer gel dosimeters are assessed in full 3D. In the intra-batch study, eight spherical flasks were filled with the same polymer gel along with a set of test tubes that served as calibration phantoms. In the inter-batch study, the eight spherical flasks were filled with different batches of gel. For each spherical phantom, a separate set of calibration phantoms was used. The spherical phantoms were irradiated using a three-field coplanar beam configuration in a very reproducible manner. The calibration phantoms were irradiated to known doses to obtain a dose-R₂ calibration plot which was applied on the corresponding R₂ maps of all spherical phantoms on an individual basis. The intra-batch study showed high dosimetric precision (3.1%) notwithstanding poor accuracy (mean dose discrepancies up to 13.0%). In the inter-batch study, a similar dosimetric precision (4.3%) and accuracy (mean dose discrepancies up to 13.7%) were found. The poor dosimetric accuracy was attributed to a systematic fault that was related to the calibration method. Therefore, the dose maps were renormalized using an independent ion chamber dose measurement. It is illustrated that with this renormalization, excellent agreement between the gel measured and TPS calculated 3D dose maps is achievable: 97% and 99% of the pixels meet the 3%/3 mm criteria for the intra- and inter-batch experiments, respectively. However renormalization will result in significant dose deviations inside a realistically sized anthropomorphic phantom as will be shown in a concurrent paper.

On the validity of 3D polymer gel dosimetry: II. physico-chemical effects.

This study quantifies some major physico-chemical factors that influence the validity of MRI (PAGAT) polymer gel dosimetry: temperature history (pre-, during and post-irradiation), oxygen exposure (post-irradiation) and volumetric effects (experiment with phantom in which a small test tube is inserted). Present results confirm the effects of thermal history prior to irradiation. By exposing a polymer gel sample to a linear temperature gradient of ∼2.8 °C cm⁻¹ and following the dose deviation as a function of post-irradiation time new insights into temporal variations were added. A clear influence of the temperature treatment on the measured dose distribution is seen during the first hours post-irradiation (resulting in dose deviations up to 12%). This effect diminishes to 5% after 54 h post-irradiation. Imposing a temperature offset (maximum 6 °C for 3 h) during and following irradiation on a series of calibration phantoms results in only a small dose deviation of maximum 4%. Surprisingly, oxygen diffusing in a gel dosimeter up to 48 h post-irradiation was shown to have no effect. Volumetric effects were studied by comparing the dose distribution in a homogeneous phantom compared to the dose distribution in a phantom in which a small test tube was inserted. This study showed that the dose measured inside the test tube was closer to the ion chamber measurement in comparison to the reference phantom without test tube by almost 7%. It is demonstrated that physico-chemical effects are not the major causes for the dose discrepancies encountered in the reproducibility study discussed in the concurrent paper (Vandecasteele and De Deene 2013a Phys. Med. Biol. 58 19-42). However, it is concluded that these physico-chemical effects are important factors that should be addressed to further improve the dosimetric accuracy of 3D MRI polymer gel dosimetry.

Radio-physical properties of micelle leucodye 3D integrating gel dosimeters

Recently, novel radiochromic leucodye micelle hydrogel dosimeters were introduced in the literature. In these studies, gel measured electron depth dose profiles were compared with ion chamber depth dose data, from which it was concluded that leucocrystal violet-type dosimeters were independent of dose rate. Similar conclusions were drawn for leucomalachite green-type dosimeters, only after pre-irradiating the samples to a homogeneous radiation dose. However, in our extensive study of the radio-physical properties of leucocrystal violet- and leucomalachite green-type dosimeters, a significant dose rate dependence was found. For a dose rate variation between 50 and 400 cGy min(-1), a maximum difference of 75% was found in optical dose sensitivity for the leucomalachite green-type dosimeter. Furthermore, the measured optical dose sensitivity of the leucomalachite green-type dosimeter was four times lower than the value previously reported in the literature. For the leucocrystal violet-type dosimeter, a maximum difference in optical dose sensitivity of 55% was found between 50 and 400 cGy min(-1). A modified composition of the leucomalachite green-type dosimeter is proposed. This dosimeter is composed of gelatin, sodium dodecyl sulfate, chloroform, trichloroacetic acid and leucomalachite green. The optical dose sensitivity amounted to 4.375 × 10(-5) cm(-1) cGy(-1) (dose rate 400 cGy min(-1)). No energy dependence for photon energies between 6 and 18 MV was found. No temperature dependence during readout was found notwithstanding a temperature dependence during irradiation of 1.90 cGy °C(-1) increase on a total dose of 100 cGy. The novel gel dosimeter formulation exhibits an improved spatial stability (2.45 × 10(-7) cm(2) s(-1) (= 0.088 mm(2) h(-1))) and good water/soft tissue equivalence. Nevertheless, the novel formulation was also found to have a significant, albeit reduced, dose rate dependence, as a maximum difference of 33% was found in optical dose sensitivity when the dose rate varied between 50 and 400 cGy min(-1). By pre-irradiating the novel leucomalachite green-type dosimeter to 500 cGy, the apparent difference in dose response between 200 and 400 cGy min(-1) was eliminated, similar to earlier findings. However, a dose response difference of 38% between 50 and 200 cGy min(-1) was still measured. On the basis of these experimental results it is concluded that the leucodye micelle gel dosimeter is not yet optimal for dose verifications of high precision radiation therapy treatments. This study, however, indicates that the dose rate dependence has a potential for improvement. Future research is necessary to further minimize the dose rate dependence through extensive chemical analysis and optimization of the gel formulation. Some insights into the physicochemical mechanisms were obtained and are discussed in this paper.

19F MRI oximetry: simulation of perfluorocarbon distribution impact

In (19)F MRI oximetry, a method used to image tumour hypoxia, perfluorocarbons serve as oxygenation markers. The goal of this study is to evaluate the impact of perfluorocarbon distribution and concentration in (19)F MRI oximetry through a computer simulation. The simulation studies the correspondence between (19)F measured (pO(FNMR)(2)) and actual tissue oxygen tension (pO(2)) for several tissue perfluorocarbon distributions. For this, a Krogh tissue model is implemented which incorporates the presence of perfluorocarbons in blood and tissue. That is, in tissue the perfluorocarbons are distributed homogeneously according to Gaussian diffusion profiles, or the perfluorocarbons are concentrated in the capillary wall. Using these distributions, the oxygen tension in the simulation volume is calculated. The simulated mean oxygen tension is then compared with pO(FNMR)(2), the (19)F MRI-based measure of pO(2) and with pO(0)(2), pO(2) in the absence of perfluorocarbons. The agreement between pO(FNMR)(2) and actual pO(2) is influenced by vascular density and perfluorocarbon distribution. The presence of perfluorocarbons generally gives rise to a pO(2) increase in tissue. This effect is enhanced when perfluorocarbons are also present in blood. Only the homogeneous perfluorocarbon distribution in tissue with no perfluorocarbons in blood guarantees small deviations of pO(FNMR)(2) from pO(2). Hence, perfluorocarbon distribution in tissue and blood has a serious impact on the reliability of (19)F MRI-based measures of oxygen tension. In addition, the presence of perfluorocarbons influences the actual oxygen tension. This finding may be of great importance for further development of (19)F MRI oximetry.

An oxygen-consuming phantom simulating perfused tissue to explore oxygen dynamics and 19F MRI oximetry

ObjectiveThis study presents a reproducible phantom which mimics oxygen-consuming tissue and can be used for the validation of 19F MRI oximetry.Materials and methodsThe phantom consists of a haemodialysis filter of which the outer compartment is filled with a gelatin matrix containing viable yeast cells. Perfluorocarbon emulsions can be added to the gelatin matrix to simulate sequestered perfluorocarbons. A blood-substituting perfluorocarbon fluid is pumped through the lumen of the fibres in the filter. 19F relaxometry MRI is performed with a fast 2D Look-Locker imaging sequence on a clinical 3T scanner.ResultsAcute and perfusion-related hypoxia were simulated and imaged spatially and temporally using the phantom.ConclusionsThe presented experimental setup can be used to simulate oxygen consumption by somatic cells in vivo and for validating computational biophysical models of hypoxia, as measured with 19F MRI oximetry.

On the reliability of 3D gel dosimetry

Gel dosimetry has a unique role to play in safeguarding conformal radiotherapy treatments as it covers the whole treatment chain and provides the radiation oncologist with the integrated dose distribution in 3D. A major obstacle that has hindered the wider dissemination of polymer gel dosimetry in radiotherapy centres is the lack of confidence in the reliability of the measured dose. Discrepancies in dose response of small versus large polymer gel dosimeters have been reported and although several hypothesis for these discrepancies have been postulated, the actual contribution of these error sources to the overall inaccuracy of the dose maps has not been determined. Several gel dosimetry research groups have chosen to use an internal calibration of gel dosimeters. In this study, the inter-and intra-batch reproducibility of the current state-of-the-art 3D gel dosimeters has been assessed. It is demonstrated that with a carefully designed scanning set-up, the overall accuracy that can be obtained with an independent calibration is well within 5% of all pixels.

Optimization of a fast optical CT scanner for nPAG gel dosimetry

A fast laser scanning optical CT scanner was constructed and optimized at the Ghent university. The first images acquired were contaminated with several imaging artifacts. The origins of the artifacts were investigated. Performance characteristics of different components were measured such as the laser spot size, light attenuation by the lenses and the dynamic range of the photo-detector. The need for a differential measurement using a second photo-detector was investigated. Post processing strategies to compensate for hardware related errors were developed. Drift of the laser and of the detector was negligible. Incorrectly refractive index matching was dealt with by developing an automated matching process. When scratches on the water bath and phantom container are present, these pose a post processing challenge to eliminate the resulting artifacts from the reconstructed images Secondary laser spots due to multiple reflections need to be further investigated. The time delay in the control of the galvanometer and detector was dealt with using black strips that serve as markers of the projection position. Still some residual ringing artifacts are present. Several small volumetric test phantoms were constructed to obtain an overall picture of the accuracy.

Low-density polymer gel dosimeters for 3D radiation dosimetry in the thoracic region: A preliminary study

Different low-density polymer gel dosimeters have been constructed that can be used to acquire the radiation dose distribution of IMRT treatments in the thoracic region. A heterogeneous phantom consisting of a low density polymer gel dosimeter sandwiched between two layers of soft tissue equivalent gel has been constructed. As a proof-of-principle, the phantom has been irradiated with a square 4 cm × 4 cm beam. The dose distribution is read out by use of both quantitative NMR spin-spin (R2) and magnetization transfer (MT) imaging. Sources of error in the dose readout have been assessed and are discussed.