Rapid Estimation of SF Parameter of Cells Using the Monte Carlo Simulation of Proton Therapy in Combination with Cascade Feed-forward Neural Network (CFNN)with LM Learning Algorithm

Please enable javascript in your browser.

Barati, Roya | 2022

149 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 55334 (46)
University: Sharif University of Technology
Department: Energy Engineering
Advisor(s): Vosoughi, Naser
Abstract:
In radiation therapy, different radiation were used to cancer treatment. The use of hadrons is more common than gamma rays for cancer treatment. Hadron therapy is appropriate to treat cancers that are close to organs at risk and can be severely damaged by X-rays during treatment. The survival fraction of cells was irradiated is important during treatment. In this study, using Geant code, simulation of choroidal melanoma tumor in the phantom of the eye was performed. The linear energy transfer and dose data of proton beam radiation were calculated, in addition, the cell survival fraction were calculated using Survival codeand the Geant simulation data such as number of inputs and dose with cell survival fraction parameter data was used to train the neural network. The results show the tumor survival parameter can be stimated by using a neural network without the need for time-consuming Monte Carlo codes such as Geant and by having inputs of the number beams and the computational survival model.
Keywords:
Proton Therapy ; GEANT4 Toolkit ; Neural Network ; Cell Survival Fraction (SF) ; Melanoma Cancer ; Radiotherapy ; Bragg Peak Curve ; Levenberg-Marquardt Equation

No TOC