SU-FF-T-333: Monte Carlo Simulations Using Whole-Body Pediatric and Adult Phantoms as Virtual Patients to Assess Secondary Organ Doses in Proton Radiation Therapy

The Monte Carlo method was employed to simulate realistic treatment situations for photon and proton radiation therapy for a set of Oak Ridge National Laboratory (ORNL) pediatric phantoms for 15, 10, 5 and 1-year olds as well as newborns. Complete radiotherapy situations were simulated using the previously developed NRUrad input code for Monte Carlo N-Particle (MCNP) code package. Each pediatric phantom was irradiated at five different positions, namely, the testes, colon, liver, left lung and brain, and the doses in targeted organs (Dt) were determined using the track length estimate of energy. The dispersed photon and proton doses in non-targeted organs (Dd), namely, the skeleton, skin, brain, spine, left and right lungs were computed. The conversion coefficients (F = Dd/Dt) of the dispersed doses were used to study the dose dispersion in different non-targeted organs for phantoms for 15, 10, 5 and 1-year olds as well as newborns. In general, the F values were larger for younger patients. The F values for non-targeted organs for phantoms for 1-year olds and newborns were significantly larger compared to those for other phantoms. The dispersed doses from proton radiation therapy were also found to be significantly lower than those from conventional photon radiation therapy. For example, the largest F values for the brain were 65.6% and 0.206% of the dose delivered to the left lung (P4) for newborns during photon and proton radiation therapy, respectively. The present results demonstrated that dispersion of photons and generated electrons significantly affected the absorbed doses in non-targeted organs during pediatric photon therapy, and illustrated that proton therapy could in general bring benefits for treatment of pediatric cancer patients.

Download Full-text

Learning-Based Stopping Power Mapping on Dual-Energy CT for Proton Radiation Therapy

International Journal of Particle Therapy ◽

10.14338/ijpt-d-20-00020.1 ◽

2021 ◽

Vol 7 (3) ◽

pp. 46-60

Author(s):

Tonghe Wang ◽

Yang Lei ◽

Joseph Harms ◽

Beth Ghavidel ◽

Liyong Lin ◽

...

Keyword(s):

Radiation Therapy ◽

Ground Truth ◽

Dual Energy ◽

Stopping Power ◽

Whole Body ◽

Chemical Compositions ◽

Proton Radiation ◽

Dose Volume Histogram ◽

Proton Radiation Therapy ◽

Dose Volume

Abstract Purpose Dual-energy computed tomography (DECT) has been used to derive relative stopping power (RSP) maps by obtaining the energy dependence of photon interactions. The DECT-derived RSP maps could potentially be compromised by image noise levels and the severity of artifacts when using physics-based mapping techniques. This work presents a noise-robust learning-based method to predict RSP maps from DECT for proton radiation therapy. Materials and Methods The proposed method uses a residual attention cycle-consistent generative adversarial network to bring DECT-to-RSP mapping close to a 1-to-1 mapping by introducing an inverse RSP-to-DECT mapping. To evaluate the proposed method, we retrospectively investigated 20 head-and-neck cancer patients with DECT scans in proton radiation therapy simulation. Ground truth RSP values were assigned by calculation based on chemical compositions and acted as learning targets in the training process for DECT datasets; they were evaluated against results from the proposed method using a leave-one-out cross-validation strategy. Results The predicted RSP maps showed an average normalized mean square error of 2.83% across the whole body volume and an average mean error less than 3% in all volumes of interest. With additional simulated noise added in DECT datasets, the proposed method still maintained a comparable performance, while the physics-based stoichiometric method suffered degraded inaccuracy from increased noise level. The average differences from ground truth in dose volume histogram metrics for clinical target volumes were less than 0.2 Gy for D95% and Dmax with no statistical significance. Maximum difference in dose volume histogram metrics of organs at risk was around 1 Gy on average. Conclusion These results strongly indicate the high accuracy of RSP maps predicted by our machine-learning–based method and show its potential feasibility for proton treatment planning and dose calculation.

Download Full-text