Recent Publications
Enhancing image quality in fast neutron-based range verification of proton therapy using a deep learning-based prior in LM-MAP-EM reconstruction
Abstract:
This study investigates the use of list-mode (LM) maximum a posteriori (MAP) expectation maximization (EM) incorporating prior information predicted by a convolutional neural network for image reconstruction in fast neutron (FN)-based proton therapy range verification. Approach. A conditional generative adversarial network (pix2pix) was trained on progressively noisier data, where detector resolution effects were introduced gradually to simulate realistic conditions. FN data were generated using Monte Carlo simulations of an 85 MeV proton pencil beam in a computed tomography-based lung cancer patient model, with range shifts emulating weight gain and loss. The network was trained to estimate the expected two-dimensional ground truth FN production distribution from simple back-projection images. Performance was evaluated using mean squared error, structural similarity index (SSIM), and the correlation between shifts in predicted distributions and true range shifts. Main results. Our results show that pix2pix performs well on noise-free data but suffers from significant degradation when detector resolution effects are introduced. Among the LM-MAP-EM approaches tested, incorporating a mean prior estimate into the reconstruction process improved performance, with LM-MAP-EM using a mean prior estimate outperforming naïve LM maximum likelihood EM (LM-MLEM) and conventional LM-MAP-EM with a smoothing quadratic energy function in terms of SSIM. Significance. Findings suggest that deep learning techniques can enhance iterative reconstruction for range verification in proton therapy. However, the effectiveness of the model is highly dependent on data quality, limiting its robustness in high-noise scenarios.
Evaluating impact of detector arrangement and position resolution effect on a fast neutron-based range verification system for proton therapy
Abstract:
Real-time proton therapy range verification is a technique that can potentially reduce uncertainty margins around the treatment volume and enable prompt corrections during treatment, making proton therapy a safer cancer treatment modality. Imaging secondary particles resulting from proton-beam nuclear interactions with tissue serves as a means of range verification. The NOVO project recently (2023) presented a compact detector array (NOVCoDA) range verification system designed to image secondary prompt-gamma rays (PGs) and fast neutrons (FNs). The position resolution and arrangement of detector elements within the NOVCoDA influences the reconstructed particle distributions and in turn the system’s range shift detection capabilities. Through Monte-Carlo simulations, we investigate the effects of four different detector element arrangements and the utilization of optically segmented scintillator volumes within detector elements, for improved position resolution, on NOVCoDA’s range shift determination capability for proton therapy. We limit our study to the detection of FNs produced by an 85-MeV proton beam interacting within a homogeneous water phantom. Results indicate that a parallel array with detector elements oriented perpendicular to the proton beam axis and line-of-sight direction yields the highest double FN scattering efficiency, on order of 10^-6 per proton. Furthermore, optically segmented detector elements resulted in improved minimum detectable range shift, reducing required proton intensity by 30%–60% to discern a 1 mm shift.
Range verification in proton therapy with computational prompt gamma-ray spectroscopy: A Monte Carlo study
Abstract:
To exploit the full potential of proton therapy, it is essential to be able to monitor the range of the proton beam. As protons have a finite range in tissue, one must rely on the measurement of secondary particles, such as prompt gamma rays and fast neutrons, to measure the range. This work is part of the NOVO project (Next generation imaging for real-time dose verification enabling adaptive proton therapy), which proposes to use a compact detector array (NOVCoDA – NOVO Compact Detector Array), based on bar shaped organic scintillators and silicon photomultipliers for light read-out, to image prompt gamma rays and fast neutrons, simultaneously. In this work, we investigate the possibility of using the NOVCoDA for prompt gamma-ray energy spectroscopy, in conjunction with the Monte Carlo Library Least Squares Approach. This method takes the whole prompt gamma-ray spectrum into account, and the assumption is that the total spectrum is a linear sum of the elemental prompt gamma-ray spectra of each constituent. We present results from Monte Carlo simulations with Geant4, where we have used a quadratic discriminant analysis classifier to identify range shifts, and determine the accuracy of the classification. With a proton intensity of 10^8 protons, the accuracy was over 90% for all range shifts with a magnitude greater than or equal to 2 mm.
Pre-project Publications
A hybrid multi-particle approach to range assessment-based treatment verification in particle therapy
Abstract:
Particle therapy (PT) used for cancer treatment can spare healthy tissue and reduce treatment toxicity. However, full exploitation of the dosimetric advantages of PT is not yet possible due to range uncertainties, warranting development of range-monitoring techniques. This study proposes a novel range-monitoring technique introducing the yet unexplored concept of simultaneous detection and imaging of fast neutrons and prompt-gamma rays produced in beam-tissue interactions. A quasi-monolithic organic detector array is proposed, and its feasibility for detecting range shifts in the context of proton therapy is explored through Monte Carlo simulations of realistic patient models and detector resolution effects. The results indicate that range shifts of can be detected at relatively low proton intensities ( protons/spot) when spatial information obtained through imaging of both particle species are used simultaneously. This study lays the foundation for multi-particle detection and imaging systems in the context of range verification in PT.
———————————————————————————————————————————-
Characterization of organic glass scintillator bars and their potential for a hybrid neutron/gamma ray imaging system for proton radiotherapy range verification
Abstract:
For accurate and simultaneous imaging of fast neutrons (FNs) and prompt gamma rays (PGs) produced during proton therapy, the selection of a highly performant detector material is crucial. In this work, a promising candidate material known as organic glass scintillator (OGS) is characterized for this task. To this end, a precisely-timed source of neutrons and Bremsstrahlung radiation produced by the nELBE facility was used to study the light output and neutron/gamma ray pulse shape discrimination (PSD) properties of a 1 × 1 × 20 cm3 OGS bar with double-sided readout. Furthermore, the energy, timing, and depth-of-interaction (DOI) resolutions of 1 × 1 × 10 cm3 and 1 × 1 × 20 cm3 OGS and EJ-200 bars were characterized with radioactive sources. For electron-equivalent energies above 0.5 MeVee, OGS was found to have excellent PSD capabilities (figure-of-merit above 1.27), energy resolution (below 12%), coincident time resolution (below 500 ps), and DOI resolution (below 10 mm). This work establishes the data analysis methods required for hybrid FN/PG imaging using OGS, and demonstrates the materials' excellent performance for this application.
———————————————————————————————————————————-
Hybrid treatment verification based on prompt gamma rays and fast neutrons: multivariate modelling for proton range determination
Abstract:
Robust and fast in vivo treatment verification is expected to increase the clinical efficacy of proton therapy. The combined detection of prompt gamma rays and neutrons has recently been proposed for this purpose and shown to increase the monitoring accuracy. However, the potential of this technique is not fully exploited yet since the proton range reconstruction relies only on a simple landmark of the particle production distributions. Here, we apply machine learning based feature selection and multivariate modelling to improve the range reconstruction accuracy of the system in an exemplary lung cancer case in silico. We show that the mean reconstruction error of this technique is reduced by 30%–50% to a root mean squared error per spot of 0.4, 1.0, and 1.9 mm for pencil beam scanning spot intensities of 108, 107, and 106 initial protons, respectively. The best model performance is reached when combining distribution features of both gamma rays and neutrons. This confirms the advantage of hybrid gamma/neutron imaging over a single-particle approach in the presented setup and increases the potential of this system to be applied clinically for proton therapy treatment verification.
———————————————————————————————————————————-
Prompt gamma-ray spectroscopy in conjunction with the Monte Carlo Library Least Squares approach: Applications to range verification in proton therapy
Abstract:
Prompt Gamma-ray Spectroscopy (PGS) in conjunction with the Monte Carlo Library Least Squares (MCLLS) approach was investigated for the purposes of range monitoring in proton therapy through Monte Carlo simulations. Prompt gamma-rays are produced during treatment and can be correlated to the range of the proton beam in the tissue. In contrast to established approaches, MCLLS does not rely on the identification of specific photopeaks. Instead it treats each individual constituent as a library spectrum and calculates coefficients for each spectrum, and therefore takes both the photopeaks and the Compton continuum into account. It can thus be applied to organic scintillators traditionally not used for energy spectroscopy due to their low Z number and density. Preliminary results demonstrate that the proposed approach returns a strong linear correlation between the range of the primary proton beam and the calculated library coefficients, depending on the composition of libraries. This can be exploited for range monitoring.