Feasibility of deformable image registration algorithm for interfractional dose calculation in radiotherapy for prostate cancer(PDF)
《中国医学物理学杂志》[ISSN:1005-202X/CN:44-1351/R]
- Issue:
- 2019年第9期
- Page:
- 995-1000
- Research Field:
- 医学放射物理
- Publishing date:
Info
- Title:
- Feasibility of deformable image registration algorithm for interfractional dose calculation in radiotherapy for prostate cancer
- Author(s):
- ZHANG Xiangbin1; 2; LI Guangjun1; ZHANG Yingjie1; BAI Sen1; 2
- 1. Department of Radiotherapy, West China Hospital of Sichuan University, Chengdu 610041, China; 2. College of Physical Science and Technology, Sichuan University, Chengdu 610065, China
- Keywords:
- ; Keywords: prostate cancer; cone beam computed tomography; deformable image registration; adaptive radiotherapy
- PACS:
- R318；R811.1
- DOI:
- DOI:10.3969/j.issn.1005-202X.2019.09.001
- Abstract:
- Abstract: Objective To evaluate the feasibility of interfractional dose evaluation based on deformable image registration (DIR) algorithm in the radiotherapy for prostate cancer. Methods A deformed cone beam CT (CBCT) with obvious registration errors was generated by the DIR from cone beam CT (CBCT) to replan CT. Based on the rigid pelvic phantom-based noise model, peak spectrum conversion method and layer-by-layer linear transformation method were used to correct the deformed CBCT and then generate two sets of virtual CBCT in which the anatomical structures were consistent with those in replan CT. The two sets of virtual CBCT were taken as the golden standard of DIR evaluation. The replan CT was taken as the baseline to evaluate the Gamma passing rate of virtual CT generated by the DIR from planned CT to virtual CBCT. The evaluation coefficients of two groups of virtual CT were compared, so as to verify the feasibility of converting replan CT to virtual CBCT. Moreover, the dice coefficient of region of interest generated by deformation field mapping was evaluated, so as to verify the feasibility of semi-automatic segmentation based on the deformation field in DIR. Results The Gamma passing rates (1 mm/1% and 2 mm/2%) of virtual CT derived from two different set of virtual CBCT were（98.08±0.43）%/（97.95±0.37）% （P>0.05） and （99.53±0.08）%/（99.55±0.10）%（P>0.05）. The dice similarity coefficient of region of interest generated based on deformation field was 0.78-0.95. Conclusion Virtual CBCT is a promising method to solve the problem that DIR algorithm lacks of golden standard and to verify the feasibility of DIR-based algorithm to evaluate the interfractional dose in the radiotherapy for prostate cancer.
Last Update: 2019-09-23