Physical parameter optimization scheme for radiobiological studies of charged particle therapy

Abstract

We have developed an easy-to-implement method to optimize the spatial distribution of a desired physical quantity for charged particle therapy. The basic methodology requires finding the optimal solutions for the weights of the constituent particle beams that together form the desired spatial distribution of the specified physical quantity, e.g., dose or dose-averaged linear energy transfer (LETd), within the target region. We selected proton, 4He ion, and 12C ion beams to demonstrate the feasibility and flexibility of our method. The pristine dose Bragg curves in water for all ion beams and the LETd for proton beams were generated from Geant4 Monte Carlo simulations. The optimization algorithms were implemented using the Python programming language. High-accuracy optimization results of the spatial distribution of the desired physical quantity were then obtained for different cases. The relative difference between the real value and the expected value of a given quantity was approximately within ± 1.0% in the whole target region. The optimization examples include a flat dose spread-out Bragg peak (SOBP) for the three selected ions, an upslope dose SOBP for protons, and a downslope dose SOBP for protons. The relative difference was approximately within ± 2.0% for the case with a flat LETd (target value = 4 keV/μm) distribution for protons. These one-dimensional optimization algorithms can be extended to two or three dimensions if the corresponding physical data are available. In addition, this physical quantity optimization strategy can be conveniently extended to encompass biological dose optimization if appropriate biophysical models are invoked.

Exploring the advantages of intensity-modulated proton therapy: experimental validation of biological effects using two different beam intensity-modulation patterns

“…In total, 94 energies of scanned proton beams are available at The University of Texas MD Anderson Cancer Center Proton Therapy Center. The dose and LET d distributions in water for each beamlet have been pre-calculated using Monte Carlo simulations 30,31 . An in-house dose optimization algorithm has been developed using the Python programming language to generate the desired dose distribution profiles 30,31 .…”

“…The dose and LET d distributions in water for each beamlet have been pre-calculated using Monte Carlo simulations 30,31 . An in-house dose optimization algorithm has been developed using the Python programming language to generate the desired dose distribution profiles 30,31 . Figure 1A shows a flat SOBP (the target region of 5.8 to 9.8 cm) and the dose contributions from its constituent beams.…”

“…distributions of all of the 94 scanned proton beams at The University of Texas MD Anderson Cancer Center Proton Therapy Center have been pre-calculated using the Monte Carlo simulation toolkit Geant4 30,31 . It should be noted that the LET d results in our simulations do not take into account the contributions of nuclear fragments and recoil nuclei from the hadronic physics processes.…”

“…The uniform dose area is 17 cm × 17 cm, large enough to cover two 96-well plates. A dose optimization algorithm for IMPT has been developed by our team using the Python programming language to generate the desired dose distribution for radiobiological studies 30,31 . In the present study, one 4-cm wide flat SOBP and one downslope SOBP were generated ranging from 5.8 to 9.8 cm in water, as shown in Fig.…”

“…In the present study, LET d is only used to "qualitatively" interpret experimental results. Our previous Monte Carlo calculation results show that at the end of range, the maximum LET d of a low-energy incident beam is higher than that of a high-energy beam 30,31 . This indicates that at the end of range, a low-energy incident beam may be more biologically effective than a high-energy beam.…”