Showing 1–2 of 2 results for author: Rahman, A U

New approaches to particle induced prompt gamma imaging

Authors: Mythra Varun Nemallapudi, Atiq Ur Rahman, Vanny Marantha Sihotang, Thien Cong Tran, Hsin-Hon Lin, Chih-Hsun Lin, Ming-Lee Chu, Augustine Ei-Fong Chen, Shih-Chang Lee

Abstract: The distribution of the prompt gamma emissions induced by proton interactions in a target material carries information concerning the proton range and target composition. In order to image prompt gamma we present different approaches. The first approach is based on knife-edge collimator utilized in tandem with a compact PET type detector in order to verify proton range by measuring the depth distr… ▽ More The distribution of the prompt gamma emissions induced by proton interactions in a target material carries information concerning the proton range and target composition. In order to image prompt gamma we present different approaches. The first approach is based on knife-edge collimator utilized in tandem with a compact PET type detector in order to verify proton range by measuring the depth distribution of prompt gamma relying on Compton scattered events. A second approach aims at unambiguously identifying the presence of protons in a region of interest by introducing a bio-compatible fiducial marker in the patient and measuring the characteristic prompt gamma of the marker against background material. A third approach illustrates a novel design for position sensitive gamma imaging with a millimeter level adjustable position resolution with a high detection efficiency which is a step towards 3D prompt gamma imaging, useful to better understand the target composition. Simulations of the designs were performed using GATE/Geant4 Monte Carlo framework. Experimental tests on fiducial marker were conducted at the proton facilities in INER and Chang Gung Memorial hospital. Using the knife-edge PET approach a range verification of 0.7 mm can be achieved for shifts within 1 cm near the tumor region and up to 4 mm for shifts within a 4 cm window. For the fiducial marker approach we identify 984 keV as the dominant prompt gamma and show that for the 50% relative PG intensity, the R80 position falls within a 3 mm thick 48Ti marker. Using the position sensitive gamma imaging approach we present a feasible design to achieve spatial resolution values of 2.6 mm with a detection efficiency of 5.4E-6 at 6.1 MeV and up to eight depth positions in a single run. The relative advantage of these methods and the challenges in the implementation are discussed. △ Less

Submitted 25 July, 2022; v1 submitted 30 April, 2022; originally announced May 2022.

Direct mapping from PET coincidence data to proton-dose and positron activity using a deep learning approach

Authors: Atiq. Ur. Rahman, Mythra Varun. Nemallapudi, Cheng-Ying. Chou, Shih-Chang Lee, Chih-Hsun. Lin

Abstract: $Objective$. Obtaining the intrinsic dose distributions in particle therapy is a challenging problem that needs to be addressed by imaging algorithms to take advantage of secondary particle detectors. In this work, we investigate the utility of deep learning methods for achieving direct mapping from detector data to the intrinsic dose distribution. $Approach… ▽ More $Objective$. Obtaining the intrinsic dose distributions in particle therapy is a challenging problem that needs to be addressed by imaging algorithms to take advantage of secondary particle detectors. In this work, we investigate the utility of deep learning methods for achieving direct mapping from detector data to the intrinsic dose distribution. $Approach$. We performed Monte Carlo simulations using GATE/Geant4 10.4 simulation toolkits to generate a dataset using human CT phantom irradiated with high-energy protons and imaged with compact in-beam PET for realistic beam delivery in a single-fraction ($\sim$2Gy). We developed a neural network model based on conditional generative adversarial networks to generate dose maps conditioned on coincidence distributions in the detector. The model performance is evaluated by the mean relative error, absolute dose fraction difference, and shift in Bragg peak position. $Main$ $results$. The relative deviation in the dose and range of the distributions predicted by the model from the true values for mono-energetic irradiation between 50 MeV and 122 MeV lie within 1% and 2%, respectively. This was achieved using $\mathrm{10^5}$ coincidences acquired five minutes after irradiation. The relative deviation in the dose and range for spread-out Bragg peak distributions were within 1% and 2.6% uncertainties, respectively. $Significance$. An important aspect of this study is the demonstration of a method for direct mapping from detector counts to dose domain using the low count data of compact detectors suited for practical implementation in particle therapy. Including additional prior information in the future can further expand the scope of our model and also extend its application to other areas of medical imaging. △ Less

Submitted 14 August, 2022; v1 submitted 29 April, 2022; originally announced April 2022.