Dosimetric evaluation of a treatment planning system using the AAPM Medical Physics Practice Guideline 5.a (MPPG 5.a) validation tests.

Verifying the accuracy of the dose calculation algorithm is considered one of the most critical steps in radiotherapy treatment for delivering an accurate dose to the patient. This work aimed to evaluate the dosimetric performance of the treatment planning system (TPS) algorithms; the AAA (v. 15.6), AXB (v. 15.6) and eMC (v. 15.6) following the AAPM medical physics practice guideline 5.a (MPPG 5.a) validation tests package in a Varian iX Linear Accelerator (Linac). A series of tests were developed based on the MPPG 5.a. on a Varian's Eclipse TPS (v. 15.6) (Varian Medical Systems). First, the basic photon and electron tests were validated by comparing the TPS calculated dose with the measurements. Next, for heterogeneity tests, we verified the Computed Tomography number to electron density (CT-to-ED) curve by comparing it with the baseline values, and TPS calculated point doses beyond heterogeneous media were compared to the measurements. Finally, for IMRT/VMAT dose validation tests, clinical reference plans were re-calculated on ArcCheck's virtual phantom (Sun Nuclear Corporation, Melbourne, FL, USA) and exported to the Linac for delivery using the ArcCheck dosimetry system. All validation tests were evaluated following the MPPG 5.a recommended tolerances. In basic dose validation tests, the TPS calculated depth dose profiles agreed well with the measurements, with a minimum gamma passing rate of 95% at 2%/2 mm criteria. However, disagreements are seen in the build-up and penumbra region. Results for most point doses in homogeneous water phantoms were within the MPPG 5.a tolerance. For the heterogeneity tests, the CT-to-ED curve was established, and calculated point doses were all within 3% of the measurements for heterogeneous media for both photon algorithms at three energies. These results are within the MPPG5.a the recommended tolerance of 3%. Moreover, for electron beams, the differences between the calculated and measured point doses averaged 5% and 7%, but were just within the MPPG 5.a tolerance of 7%. For IMRT and VMAT validation tests using a gamma criteria of a 2%/2 mm, IMRT plans showed maximum and minimum passing rates of 98.2% and 97.4%, respectively. Whereas VMAT plans showed maximum and minimum passing rates of 100% and 94.3%, respectively. We conclude that the dosimetric accuracy of the Eclipse TPS (v15.6) algorithm is adequate for clinical use. The MPPG 5.a tests are valuable for evaluating dose calculation accuracy and are very useful for TPS upgrade checks, commissioning tests, and routine TPS QA.

An evaluation of solid state detectors for the relative dosimetry of kilovoltage X-ray beams.

INTRODUCTION: Kilovoltage (kV) X-ray beams are an essential modality in radiotherapy. Solid state detectors are widely available in radiotherapy departments, but their use for kV dosimetry has been limited to date. This study aimed to evaluate the dosimetric performance of a range of solid state detectors for kV dosimetry. METHOD: Percentage depth doses (PDDs) and relative output factors (ROFs) were measured on an XStrahl 300 unit (XStrahl-Ltd., UK) using 60, 100, 150, and 300 kVp X-ray beams. The fields were defined by circular applicators with field sizes of 2, 5, 8, and 10 cm diameter and square applicators of field sizes 10 × 10 and 20 × 20 cm2 . The following Physikalisch-Technische Werkstätten (PTW) dosimeters were used for measurements: Advanced Markus, PinPoint 3D and Semiflex ionization chambers; photon, electron, and stereotactic radiosurgery (SRS) diodes plus the microDiamond detector. All PDDs were normalized at 5 mm depth, and ROFs were measured at 3 mm depth to avoid collisions with the end of the applicators. ROFs measured using chambers were corrected for polarity and ion-recombination effects. RESULTS AND DISCUSSION: PDD measurements for 60, 100, and 150 kVp beams exhibited good agreement between all diodes and the ionization chambers over the entire range of depths except in the first few millimeters near the surface. However, for the 300 kVp, all diode detectors exhibited an overresponding behavior compared to reference depth dose data measured with the Advanced Markus chamber. ROFs with the diodes were higher than the Advanced Markus chamber at low energy, and the magnitude of these differences is inversely proportional to the field sizes. The PTW P diode showed the highest variation of up to 15% in the output factor compared to the Advanced Markus chamber. CONCLUSION: This study evaluated the dosimetric performance of a range of solid state detectors in kV relative dosimetry. This study showed that diode detectors are a suitable replacement for ionization chambers for the PDD measurement of low energy kV beams (60-150 kVp) except for the PDD of 60 kVp with the smaller field sizes. However, an overresponding behavior of diode detectors at 300 kVp beams shows that diode detectors are not suitable for the PDD measurement of high energy kV beams. Generally, all solid state detectors overresponded to ROF measurements, indicating that it is not suitable for ROF measurements. In general, both shielded and unshielded diodes produced a similar dosimetric response, which demonstrates that the energy dependence of solid state detectors should be considered before they are used for any kV relative dosimetric measurements.

From AAA to Acuros XB-clinical implications of selecting either Acuros XB dose-to-water or dose-to-medium.

When implementing Acuros XB (AXB) as a substitute for anisotropic analytic algorithm (AAA) in the Eclipse Treatment Planning System, one is faced with a dilemma of reporting either dose to medium, AXB-Dm or dose to water, AXB-Dw. To assist with decision making on selecting either AXB-Dm or AXB-Dw for dose reporting, a retrospective study of treated patients for head & neck (H&N), prostate, breast and lung is presented. Ten patients, previously treated using AAA plans, were selected for each site and re-planned with AXB-Dm and AXB-Dw. Re-planning was done with fixed monitor units (MU) as well as non-fixed MUs. Dose volume histograms (DVH) of targets and organs at risk (OAR), were analyzed in conjunction with ICRU-83 recommended dose reporting metrics. Additionally, comparisons of plan homogeneity indices (HI) and MUs were done to further highlight the differences between the algorithms. Results showed that, on average AAA overestimated dose to the target volume and OARs by less than 2.0 %. Comparisons between AXB-Dw and AXB-Dm, for all sites, also showed overall dose differences to be small (<1.5 %). However, in non-water biological media, dose differences between AXB-Dw and AXB-Dm, as large as 4.6 % were observed. AXB-Dw also tended to have unexpectedly high 3D maximum dose values (>135 % of prescription dose) for target volumes with high density materials. Homogeneity indices showed that AAA planning and optimization templates would need to be adjusted only for the H&N and Lung sites. MU comparison showed insignificant differences between AXB-Dw relative to AAA and between AXB-Dw relative to AXB-Dm. However AXB-Dm MUs relative to AAA, showed an average difference of about 1.3 % signifying an underdosage by AAA. In conclusion, when dose is reported as AXB-Dw, the effect that high density structures in the PTV has on the dose distribution should be carefully considered. As the results show overall small dose differences between the algorithms, when transitioning from AAA to AXB, no significant change to existing prescription protocols is expected. As most of the clinical experience is dose-to-water based and calibration protocols and clinical trials are also dose-to-water based and there still exists uncertainties in converting CT number to medium, selecting AXB-Dw is strongly recommended.

National dosimetric audit network finds discrepancies in AAA lung inhomogeneity corrections.

This work presents the Australian Clinical Dosimetry Service's (ACDS) findings of an investigation of systematic discrepancies between treatment planning system (TPS) calculated and measured audit doses. Specifically, a comparison between the Anisotropic Analytic Algorithm (AAA) and other common dose-calculation algorithms in regions downstream (≥2cm) from low-density material in anthropomorphic and slab phantom geometries is presented. Two measurement setups involving rectilinear slab-phantoms (ACDS Level II audit) and anthropomorphic geometries (ACDS Level III audit) were used in conjunction with ion chamber (planar 2D array and Farmer-type) measurements. Measured doses were compared to calculated doses for a variety of cases, with and without the presence of inhomogeneities and beam-modifiers in 71 audits. Results demonstrate a systematic AAA underdose with an average discrepancy of 2.9 ± 1.2% when the AAA algorithm is implemented in regions distal from lung-tissue interfaces, when lateral beams are used with anthropomorphic phantoms. This systemic discrepancy was found for all Level III audits of facilities using the AAA algorithm. This discrepancy is not seen when identical measurements are compared for other common dose-calculation algorithms (average discrepancy -0.4 ± 1.7%), including the Acuros XB algorithm also available with the Eclipse TPS. For slab phantom geometries (Level II audits), with similar measurement points downstream from inhomogeneities this discrepancy is also not seen.