Ponente
Descripción
Proton-range verification is an important challenge in proton radiotherapy. Many methods have been proposed to reduce the uncertainty in the localization of the deposited dose with these treatments. The radio-induced thermoacoustic effect (i.e. the conversion of some of the deposited energy in a tissue into acoustic waves) can be used to measure the penetration depth of the proton beams in real-time. This method has significant advantages compared to other alternatives, as it requires a low-cost and small equipment, but it is challenging due to the intrinsic low signal-to-noise ratio (SNR) of the measured data and the complex propagation of the acoustic waves in heterogeneous media. In this work, we present several algorithms and regularization methods for protoacoustic image reconstruction, and evaluated which one is able to provide better image quality with very noisy data. We used simulated data of the deposited dose of a proton beam in a water tank, converted into an initial pressure-wave using the dose-acoustic equation, and then propagated the acoustic wave in the medium using the software k-Wave. We finally added zero-mean Gaussian noise to the resulting signal recorded by a transducer placed in the beam direction. Our results indicate that even with noisy data, both the gradient-descent and a adapted version of the MLEM algorithm, commonly used in other medical imaging techniques such as Positron Emission Tomography, can be used to successfully reconstruct the dose distribution. These promising results have to be validated with real data acquired in a proton-beam facility.
