Room: Exhibit Hall | Forum 4
Purpose: Treatment accuracy of proton therapy is limited due to proton range uncertainty, which may result in an overshoot or undershoot of target volume. The objective of this study is to reconstruct dose distribution and localize Bragg peak using proton-induced acoustic signals within patients.
Methods: A 3D heterogeneous patient phantom was built by assembling multiple consecutive diagnostic CT images. CT number in each voxel was converted into tissue parameters, including mass density (Ï?), speed of sound (c), thermal expansion (Î²), heat capacity (Cp), acoustic attenuation (Î±) and elemental weight (w). A number of transducers locating on the surface of an object are used to detect acoustic signals. Based on a k-space pseudospectral finite element method, a time reversal based protoacoustic image reconstruction was performed. Initial protoacoustic distribution generated by the energy dose deposition was modeled by Monte Carlo simulation based on tissue properties. The k-Wave Matlab toolbox was used to simulate acoustic propagation using transducers. The method was evaluated using an abdomen digital phantom.
Results: The proposed method is able to identify the location of Bragg peak within 1 mm (pixel size: 0.59 mm), when the noise of transducers is not considered. The results also suggest that quality of reconstructed dose distribution depends on three factors: number of sensors, sampling interval, and iterative algorithm.
Conclusion: The proposed method is a promising tool to be used for range and dose verification in proton therapy. Compared to other online monitoring solution, it offers several unique advantages including online monitor, positioning accuracy, dose recovery and cost-effectiveness.
Protons, Image Analysis, Ultrasonics
Not Applicable / None Entered.