Accuracy verification of GPU-accelerated collapsed cone convolution algorithm for dose calculation(PDF)
《中国医学物理学杂志》[ISSN:1005-202X/CN:44-1351/R]
- Issue:
- 2023年第9期
- Page:
- 1075-1082
- Research Field:
- 医学放射物理
- Publishing date:
Info
- Title:
- Accuracy verification of GPU-accelerated collapsed cone convolution algorithm for dose calculation
- Author(s):
- L?Ang1; 2; XU Zihao1; 2; SUN Guangyao3; ZHAO Yun3; ZHENG Huaqing3; 4; 5; JIA Jing3; 4; 5; JIANG Jieqiong1
- 1. Hefei Institutes of Physical Science, Chinese Academy of Sciences, Hefei 230031, China 2. Science Island Branch of Graduate School, University of Science and Technology of China, Hefei 230031, China 3. International Academy of Neutron Science, Qingdao 266199, China 4. SuperAccuracy Science & Technology Co., Ltd, Nanjing 210044, China 5. SuperAccuracy Advanced Technology Research Institute Co., Ltd, Hefei 230031, China
- Keywords:
- Keywords: collapsed cone convolution treatment planning system photon radiotherapy dose calculation
- PACS:
- R318；R811.1
- DOI:
- DOI:10.3969/j.issn.1005-202X.2023.09.004
- Abstract:
- Abstract: Objective To verify the accuracy and clinical feasibility of GPU-accelerated collapsed cone convolution algorithm for dose calculation. Methods The GPU-accelerated collapsed cone convolution algorithm was tested based on the national standards YY 0775, YY/T 0889 and some clinical cases with the treatment planning system KylinRay-TPSe as platform. In the test by the national standard YY 0775, the dose at a specified calculation point in a water phantom under specific irradiation geometrical condition was calculated and compared with a reference dose. In the test by the national standard YY/T 0889, the dose at a specific point or plane on the solid water phantom in intensity-modulated radiotherapy plan was calculated and compared with the dose value measured in the experiment. The clinical case test was carried out on 5 clinical cases of head and neck tumor, 5 of chest and abdomen tumor and 5 of pelvic tumor. The patients treatment plan information was exported from the Pinnacle3 TPS and imported into KylinRay-TPSe, and the dose distribution was calculated with GPU-accelerated collapsed cone convolution algorithm. The dosimetric parameters of target areas and organs-at-risk were compared with the calculation results of Pinnacle3 TPS by t-test, and the 2D gamma passing rates of the isocenter plane and the 3D gamma passing rates were compared. Results The GPU-accelerated collapsed cone convolution dose algorithm met the accuracy requirements of the national standards YY 0775 and YY/T 0889. In the clinical case test, there was no statistically significant difference in the dosimetric parameters of target areas and organs-at-risk ([P>0.05]), and the 2D gamma passing rates of the isocenter plane and the 3D gamma passing rates were above 95.1% and 95.0%. Conclusion The GPU-accelerated collapsed cone convolution algorithm has high dose calculation accuracy and meets the requirements for clinical treatment.
Last Update: 2023-09-26