ESTRO 38 Abstract book

S959 ESTRO 38

values interpolated from the DVH text file agreed within 1% with the values calculated by the TPS. The program was also found to be sensitive to errors created deliberately in dummy cases such as incorrect dose prescription, inconsistent catheter lengths and dwell positions and wrong reference source data. The process of exporting the files, running the program and creating the reports can be completed in less than 5 minutes. Conclusion The program is a useful tool to independently verify the dose calculation in brachytherapy treatments and had led to a more efficient workflow in the clinical sessions. The automated EQD2 report has reduced the possibility of transcription errors. The use of DICOM and DVH text files offers the possibility of extending the program to other TPS. EP-1775 Determination of tolerance criteria for the sliding leaf gap dynamic IMRT test S. Weston 1 , C. Thompson 1 , A. Esmail 1 , P. Rixham 1 , D. Paynter 1 1 Leeds Cancer Centre, Medical Physics and Engineering, Leeds, United Kingdom Purpose or Objective In the sliding leaf gap test a dynamic narrow field is moved across an ion chamber and the charge collected is normalised by an open field delivered using the same number of monitor units. This work determines clinically relevant tolerance criteria by introducing MLC errors into the delivery, and comparing VMAT verification results performed using Delta4 and ion chamber measurements within a phantom with sliding window output factor (SWOF) measurement. Material and Methods The MLC leaves on the LINAC were first calibrated and QA'd to ensure that the MLCs were performing optimally. The accuracy of treatment delivery was assessed by a measurement using Delta4 and by an ion chamber measurement in a phantom at the isocentre. Each leaf bank was adjusted by modifying the MLC calibration such that the subfields were made larger by 0.1 mm on each leaf. The verification measurements were repeated. This process of modifying the leaf bank calibration and verifying the VMAT deliveries was repeated until one of the deliveries failed the verification. The MLC calibration was restored and the process repeated this time reducing the subfield size. A water phantom was placed at 100 cm SSD with an ion chamber on the CAX at depth of 5cm. The chamber was irradiated with a 200 MU 10x10 cm 2 field. The chamber was irradiated with a dynamic 200 MU 10 x 1 cm 2 field which moved across the field from -4.5 cm to 4.5 cm. The ratio of the dynamic field output to the open field is the sliding window output factor (SWOF). The SWOF was recorded for range of leaf calibration settings assessed as being clinically acceptable using the verification measurements of the two VMAT solutions. This process was repeated across 7 linear accelerators and the results from all linacs used to determine the clinical tolerances for the SWOF. Results The verification results for both anus and endometrium for LINAC A are shown in figure 1. The figure shows that as the leaf calibration is extended or withdrawn the verification results become poorer until eventually the verifications fail..

).

Material and Methods For each fraction, DICOM RP and DVH text files are exported from the TPS, OncentraBrachy v4.5. The matlab program uses the source coordinates from the RP file along with published source data (Ir-192) as the primary input for dose calculation following TG 43 protocol and accounts for 2D anisotropy. The independent dose calculation was performed on point A and bladder/rectum dose points. The results were compared to the TPS calculation. Furthermore, the program verifies other treatments parameters which includes: total reference air kerma, dose prescription, applicator type, dwell positions, catheter length, total treatment time and a comparison to known standard time if appropriate. The Matlab program also imports the tabular DVH file to interpolate the 2cc, 1cc and 0.1cc dose values of bladder, rectum and bowel and D100, D90 and V100 for the HRCTV. The program automatically exports the results to a patient report (Figure 2), and sums all doses at final fraction. Sensitivity of the program was studied by introducing deliberate errors. Results 552 dose points have been checked with the independent dose calculation since the program was implemented in clinical practice. The EQD2 report created with Matlab has been used for 30 patients. The mean error in dose calculation between the Matlab code and TPS was -0.02% ± 0.61% [-5.12, 1.51][YA1] . Largest errors were found on points created on the same plane as the ring applicator due to the different vectors used for each dwell position by the manufacturer and our measured ring path that affect the TG 43 interpolations. The OARs and HRCTV