Objective To develop a rapid and accurate Monte Carlo program(simplified code for dosimetry of carbon ions,SPEEDO)for carbon ion therapy.Methods For electromagnetic process,type Ⅱ condensed history simulation scheme and continuous slowing down approximation were used to simulate energy straggling,range straggling,multiple scattering,and ionization processes.For nuclear interaction,5 types of target nuclei were considered,including hydrogen,carbon,nitrogen,oxygen,and calcium.The produced secondary charged particles followed the same condensed history framework.The study simulated the transport of carbon ions in 4 materials(water,soft tissues,lung,and bone),and the calculated doses were validated against TOPAS(a Monte Carlo simulation software for radiotherapy physics),followed by a comparison with dose measurements in a water phantom from the HIMM-WW(a medical heavy-ion accelerator facility in Wuwei).Results SPEEDO's simulation results showed good consistency with TOPAS.For each material,in the voxel region where the physical dose was greater than 10％of the maximum dose point,the relative maximum dose error of both was less than 2％.At treatment energy of 400 MeV/u,SPEEDO's computation time was significantly less than that of TOPAS(13.8 min vs 105.0 min).SPEEDO's calculation results also showed good agreement with HIMM-WW measurements in terms of lateral dose distribution and integrated dose depth curve.Conclusion SPEEDO program can accurately and rapidly perform Monte Carlo dose calculations for carbon-ion therapy.

Objective To develop a GPU-based Monte Carlo dose calculation module for helical tomotherapy(TOMO),and integrate it into the commercial software ArcherQA to achieve fast and accurate dose verification in clinic.Methods The TOMO treatment head was modeled using TOPAS to obtain phase space files,and a fast weight tuning algorithm was used to simulate particle transport in multi-leaf collimator for improving computational efficiency,and finally,GPU-based Monte Carlo algorithms in ArcherQA were used to simulate particle transport in patients.To verify the model accuracy,the ArcherQA calculated results in water tank were compared with measured data for different open fields.In addition,multiple comparisons among ArcherQA results,TPS results and ArcCHECK results were conducted on 15 clinical cases(5 cases in the head and neck,5 cases in the chest and abdomen,and 5 cases in the whole body).Results In the water tank tests for 40 cm×5.0 cm,40 cm×2.5 cm and 40 cm× 1.0 cm radiation fields,the average global relative errors of the percentage depth dose,transverse dose distribution,and longitudinal dose distribution calculated by ArcherQA with the corresponding measured values were 0.72％,0.66％,and 0.54％,respectively.Over 98％of the voxels had a global relative error of less than 1％.As for 15 clinical cases,in 2％/2 mm criteria,the mean Gamma passing rate was 98.1％between ArcherQA and TPS,99.1％between TPS and ArcCHECK,and 99.4％between ArcherQA and ArcCHECK.The uncertainty of the simulation maintained less than 1％,and the average time taken for calculation based on patient CT vs ArcCHECK phantom was 87 s vs 64 s.Conclusion ArcherQA can be used for independent dose validation for TOMO plans for it can provide fast and accurate dose calculations.

Objective:To develop a deep learning model for predicting three-dimensional (3D) voxel-wise dose distributions for intensity-modulated radiotherapy (IMRT).Methods:A total of 110 postoperative rectal cancer cases treated by IMRT were considered in the study, of which 90 cases were randomly selected as the training-validating set and the remaining as the testing set. A 3D deep learning model named 3D U-Res-Net was constructed to predict 3D dose distributions. Three types of 3D matrices from CT images, structure sets and beam configurations were fed into the independent input channel, respectively, and the 3D matrix of IMRT dose distributions was taken as the output to train the 3D model. The obtained 3D model was used to predict new 3D dose distributions. The predicted accuracy was evaluated in two aspects: the average dose prediction bias and mean absolute errors (MAEs)of all voxels within the body, the dice similarity coefficients (DSCs), Hausdorff distance(HD 95) and mean surface distance (MSD) of different isodose surfaces were used to address the spatial correspondence between predicted and clinical delivered 3D dose distributions; the dosimetric index (DI) including homogeneity index, conformity index, V50, V45 for PTV and OARs between predicted and clinical truth were statistically analyzed with the paired-samples t test. Results:For the 20 testing cases, the average prediction bias ranged from -2.12% to 2.88%, and the MAEs varied from 2.55% to 5.75%. The DSCs value was above 0.9 for all isodose surfaces, the average MSD ranged from 0.21 cm to 0.45 cm, and the average HD 95 varied from 0.61 cm to 1.54 cm. There was no statistically significant difference for all DIs, except for bladder Dmean. Conclusions:This study developed a deep learning model based on 3D U-Res-Net by considering beam configurations input and achieved an accurate 3D voxel-wise dose prediction for rectal cancer treated by IMRT.