(Sunday, 7/14/2019) 4:00 PM - 4:30 PM
Room: Exhibit Hall | Forum 7
Purpose: To develop and validate a GPU-based Collapsed Cone dose calculation engine integrated in DeepPlan (a new commercial TPS) to improve the efficiency and accuracy of plan design in radiotherapy.
Methods: An improved collapsed cone dose calculation Algorithm based on quadratic integral kernels instead of the original point spread kernels, was developed which was better to deal with voxel size effects and could be more accurate for multi-resolution dose calculation. Graphics processing unit (GPU) was utilized to achieve high efficiency of dose calculation engine. We have also developed a LINAC-specific beam model method which considering the primary source, the scatter source, off-axis beam profile, MLC tongue-and-groove, MLC leaf tip width, collimator position offset etc. The beam model parameters were adapted during beam commission to provide the best fit to the measured data. Two clinical plans for head and prostate cases were investigated to compare the composite/field-by-field planar dose distribution with measurements.
Results: Speed-up factors of 3-5 for dose calculation have been observed using an NVIDIA GeForce 1060 GPU card against a 4.2 GHz Intel core i7-7700K CPU processor. The time of dose calculation for head and prostate were 5.2s and 8.7s, respectively. The 2D 3%/3mm gamma analysis for the plane dose was performed for the two clinical cases. The gamma passing rates for all 9 fields of the head tumor case were higher than 95%. For the breast case, the passing rates of all 7 fields were higher than 97%. The gamma passing rates of composite plane dose for head and prostate case were 97.9% and 100%, respectively.
Conclusion: An improved GPU-based collapsed cone dose calculation engine has been developed and integrated in DeepPlan to provide a streamlined workflow for routine treatment plan design.
Funding Support, Disclosures, and Conflict of Interest: Supported by the National Key Research and Development Program of China (No.2017YFC0107500) and the National Natural Science Foundation of China (Grant No.11575180)
Keywords
Not Applicable / None Entered.
Taxonomy
Not Applicable / None Entered.
Contact Email
