Purpose: Accurate online determining of the elemental composition of irradiated tissue in proton therapy plays an important role in treatment verification, since it leads to accurate estimation of the depth of Bragg peak, the proton range and the stopping power map of irradiated tissue. A hopeful approach for elemental analysis in proton therapy is employing the proton induce characteristic prompt gamma-rays. In this work, the effect of gamma spectrum noise level on estimating elemental analysis of irradiated target and proton range are investigated.
Methods: In the present study, the Geant4 toolkit has been used to simulate the prompt gamma-ray spectra of phantom in different noise levels. Analysis done using correlation between the mass concentration of elements and net counts of selected peaks in proton induced prompt gamma spectra of irradiated phantoms. To investigate the effect of noise levels on the accuracy and precision of the results, different noise levels are simulated, and the errors of elemental mass are calculated. Also the error propagation applied to estimate error of analytical calculated proton range.
Results: The results show that both the accuracy and the precision of estimated elemental weight percent decrease by increasing noise levels. The quantification results show that the elemental weight percent errors in high noise level lead to deviations less than 1% in mass rang (in g/cm2) of protons.
Conclusion: In conclusion, the applied analysis method can be used to estimate weight percent of irradiated tissue from its recorded prompt gamma spectra even in the presence of high noise levels. However, the increment in the standard deviation of the results is unavoidable.