Loading...

Feasibility Study of Rapid Estimation of SF Parameter of Cells Using the Monte Carlo Simulation of Carbon Therapy in Combination with Cascade Feed-forward Neural Network (CFNN) by Bayesian Regularization Learning Algorithm

Khodadadi, Rezvan | 2022

149 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 55333 (46)
  4. University: Sharif University of Technology
  5. Department: Energy Engineering
  6. Advisor(s): Vosoughi, Naser
  7. Abstract:
  8. Our main goal in this project is to calculate the cell survival fraction (SF) in unhealthy tissue after carbon ion beams are emitted and hit the target tissue. At present, in the stages of carbon therapy, it usually takes a long time to make the necessary calculations for that particular person, and this makes the treatment process progress slowly, but in this project, we intend to address this problem with Eliminating the use of neural network. The main reason is the use of carbon ions to increase efficiency in the treatment of tumors and deep cancerous tissues. To do this project, we first convert the carbon ion energy to the dose absorbed at the desired point. For this phase of the research, data on the Bragg peak curve, dose data and linear energy transfer of carbon ions in the human ocular tumor were extracted from the Geant4 simulation tool using an advanced example of Hadrontherapy. Next, the Survival analysis code is used to calculate the cell survival fraction data. By specifying input data such as the corresponding α and β gamma rays values, the dose and LET-Dose for carbon ions, cell type and radius are calculated by direct linear interpolation of the cell survival data by the Survival analysis code.Then, to speed up the calculations, due to the slow calculations related to the absorbed dose and the cell survival fraction, we train the neural network with Bayesian algorithm by considering the dose data obtained from the simulation and the data related properties. Finally, the data obtained from the network training are validated with the experimental data related to the reference articles, and the errors resulting from the comparison of these data and the evaluation of the results are given
  9. Keywords:
  10. GEANT4 Toolkit ; Bayesian Neural Networks ; Monte Carlo Method ; Bragg Cell ; Bragg Peak Curve ; Carbon Therapy ; Cell Survival Fraction (SF) ; Cancer

 Digital Object List

 Bookmark

No TOC