Dosimetric Effects on Small-field Beam-modeling for Stereotactic Body Radiation Therapy

作者：Cho Woong^*; Kim Suzy; Kim Jung In; Wu Hong Gyun; Jung Joo Young; Kim Min Joo; Suh Tae Suk; Kim Jin Young; Kim Jong Won

来源：Journal of the Korean Physical Society, 2015, 66(4): 678-693.

DOI：10.3938/jkps.66.678

• Small field dosimetry
• Beam-modeling
• Treatment planning system
• SBRT

The treatment planning of stereotactic radiosurgery (SRS) and stereotactic body radiation therapy (SBRT) requires high accuracy of dosimetric data for small radiation fields. The dosimetric effects on the beam-modeling process of a treatment planning system (TPS) were investigated using different measured small-field data sets. We performed small-field dosimetry with three detectors: a CC13 ion chamber, a CC01 ion chamber, and an edge detector. Percentage depth doses (PDDs) and dose profiles for field sizes given by 3 x 3 cm(2), 2 x 2 cm(2), and 1 x 1 cm(2) were obtained for 6 MV and 15 MV photon beams. Each measured data set was used as data input for a TPS, in which a beam-modeling process was implemented using the collapsed cone convolution (CCC) algorithm for dose calculation. The measured data were used to generate six beam-models based on each combination of detector type and photon energy, which were then used to calculate the corresponding PDDs and dose profiles for various depths and field sizes. Root mean square differences (RMSDs) between the calculated and the measured doses were evaluated for the PDDs and the dose profiles. The RMSDs of PDDs beyond the maximum dose depth were within an accuracy of 0.2 - 0.6%, being clinically acceptable. The RMSDs of the dose profiles corresponding to the CC13, the CC01, and the edge detector were 2.80%, 1.49%, and 1.46% for a beam energy of 6 MV and 2.34%, 1.15%, and 1.44% for a beam energy of 15 MV, respectively. The calculated results for the CC13 ion chamber showed the most discrepancy compared to the measured data, due to the relatively large sensitive volume of this detector. However, the calculated dose profiles for the detectors were not significantly different from another. The physical algorithm used in the beam-modeling process did not seem to be sensitive to blurred data measured with detectors with large sensitive volumes. Each beam-model was used to clinically evaluate lung and lymphatic node SBRT plans, yielding almost equal dose distributions for the treatment targets, while the mean doses related to the organs at risk (OAR,$) deviated by approximately 0.7 - 1.2%. The use of the measured data sets from different detectors for the beam-modeling process still provided acceptable dose distributions with accuracies within 2%.

• 出版日期2015-2