A feasibility study of enhanced prompt gamma imaging for range verification in proton therapy using deep learning.

Background and objective. Range uncertainty is a major concern affecting the delivery precision in proton therapy. The Compton camera (CC)-based prompt-gamma (PG) imaging is a promising technique to provide 3Din vivorange verification. However, the conventional back-projected PG images suffer from severe distortions due to the limited view of the CC, significantly limiting its clinical utility. Deep learning has demonstrated effectiveness in enhancing medical images from limited-view measurements. But different from other medical images with abundant anatomical structures, the PGs emitted along the path of a proton pencil beam take up an extremely low portion of the 3D image space, presenting both the attention and the imbalance challenge for deep learning. To solve these issues, we proposed a two-tier deep learning-based method with a novel weighted axis-projection loss to generate precise 3D PG images to achieve accurate proton range verification.Materials and methods: the proposed method consists of two models: first, a localization model is trained to define a region-of-interest (ROI) in the distorted back-projected PG image that contains the proton pencil beam; second, an enhancement model is trained to restore the true PG emissions with additional attention on the ROI. In this study, we simulated 54 proton pencil beams (energy range: 75-125 MeV, dose level: 1 × 109protons/beam and 3 × 108protons/beam) delivered at clinical dose rates (20 kMU min-1and 180 kMU min-1) in a tissue-equivalent phantom using Monte-Carlo (MC). PG detection with a CC was simulated using the MC-Plus-Detector-Effects model. Images were reconstructed using the kernel-weighted-back-projection algorithm, and were then enhanced by the proposed method.Results. The method effectively restored the 3D shape of the PG images with the proton pencil beam range clearly visible in all testing cases. Range errors were within 2 pixels (4 mm) in all directions in most cases at a higher dose level. The proposed method is fully automatic, and the enhancement takes only ∼0.26 s.Significance. Overall, this preliminary study demonstrated the feasibility of the proposed method to generate accurate 3D PG images using a deep learning framework, providing a powerful tool for high-precisionin vivorange verification of proton therapy.

Applications of Machine Learning to Improve the Clinical Viability of Compton Camera Based in vivo Range Verification in Proton Radiotherapy.

We studied the application of a deep, fully connected Neural Network (NN) to process prompt gamma (PG) data measured by a Compton camera (CC) during the delivery of clinical proton radiotherapy beams. The network identifies 1) recorded "bad" PG events arising from background noise during the measurement, and 2) the correct ordering of PG interactions in the CC to help improve the fidelity of "good" data used for image reconstruction. PG emission from a tissue-equivalent target during irradiation with a 150 MeV proton beam delivered at clinical dose rates was measured with a prototype CC. Images were reconstructed from both the raw measured data and the measured data that was further processed with a neural network (NN) trained to identify "good" and "bad" PG events and predict the ordering of individual interactions within the good PG events. We determine if NN processing of the CC data could improve the reconstructed PG images to a level in which they could provide clinically useful information about the in vivo range and range shifts of the proton beams delivered at full clinical dose rates. Results showed that a deep, fully connected NN improved the achievable contrast to noise ratio (CNR) in our images by more than a factor of 8x. This allowed the path, range, and lateral width of the clinical proton beam within a tissue equivalent target to easily be identified from the PG images, even at the highest dose rates of a 150 MeV proton beam used for clinical treatments. On average, shifts in the beam range as small as 3 mm could be identified. However, when limited by the amount of PG data measured with our prototype CC during the delivery of a single proton pencil beam (~1 × 109 protons), the uncertainty in the reconstructed PG images limited the identification of range shift to ~5 mm. Substantial improvements in CC images were obtained during clinical beam delivery through NN pre-processing of the measured PG data. We believe this shows the potential of NNs to help improve and push CC-based PG imaging toward eventual clinical application for proton RT treatment delivery verification.

Challenges and opportunities for the simulation of calcium waves on modern multi-core and many-core parallel computing platforms.

State-of-the-art distributed-memory computer clusters contain multicore CPUs with 16 and more cores. The second generation of the Intel Xeon Phi many-core processor has more than 60 cores with 16 GB of high-performance on-chip memory. We contrast the performance of the second-generation Intel Xeon Phi, code-named Knights Landing (KNL), with 68 computational cores to the latest multicore CPU Intel Skylake with 18 cores. A special-purpose code solving a system of nonlinear reaction-diffusion partial differential equations with several thousands of point sources modeled mathematically by Dirac delta distributions serves as realistic test bed. The system is discretized in space by the finite volume method and advanced by fully implicit time-stepping, with a matrix-free implementation that allows the complex model to have an extremely small memory footprint. The sample application is a seven variable model of calcium-induced calcium release (CICR) that models the interplay between electrical excitation, calcium signaling, and mechanical contraction in a heart cell. The results demonstrate that excellent parallel scalability is possible on both hardware platforms, but that modern multicore CPUs outperform the specialized many-core Intel Xeon Phi KNL architecture for a large class of problems such as systems of parabolic partial differential equations.

Merging Orthovoltage X-Ray Minibeams spare the proximal tissues while producing a solid beam at the target.

Conventional radiation therapy of brain tumors often produces cognitive deficits, particularly in children. We investigated the potential efficacy of merging Orthovoltage X-ray Minibeams (OXM). It segments the beam into an array of parallel, thin (~0.3 mm), planar beams, called minibeams, which are known from synchrotron x-ray experiments to spare tissues. Furthermore, the slight divergence of the OXM array make the individual minibeams gradually broaden, thus merging with their neighbors at a given tissue depth to produce a solid beam. In this way the proximal tissues, including the cerebral cortex, can be spared. Here we present experimental results with radiochromic films to characterize the method's dosimetry. Furthermore, we present our Monte Carlo simulation results for physical absorbed dose, and a first-order biologic model to predict tissue tolerance. In particular, a 220-kVp orthovoltage beam provides a 5-fold sharper lateral penumbra than a 6-MV x-ray beam. The method can be implemented in arc-scan, which may include volumetric-modulated arc therapy (VMAT). Finally, OXM's low beam energy makes it ideal for tumor-dose enhancement with contrast agents such as iodine or gold nanoparticles, and its low cost, portability, and small room-shielding requirements make it ideal for use in the low-and-middle-income countries.

Nuclear introns help unravel the diversification history of the Australo-Pacific Petroica robins.

Australo-Pacific Petroica robins are known for their striking variability in sexual plumage coloration. Molecular studies in recent years have revised the taxonomy of species and subspecies boundaries across the southwest Pacific and New Guinea. However, these studies have not been able to resolve phylogenetic relationships within Petroica owing to limited sampling of the nuclear genome. Here, we sequence five nuclear introns across all species for which fresh tissue was available. Nuclear loci offer support for major geographic lineages that were first inferred from mtDNA. We find almost no shared nuclear alleles between currently recognized species within the New Zealand and Australian lineages, whereas the Pacific robin radiation has many shared alleles. Multilocus coalescent species trees based on nuclear loci support a sister relationship between the Australian lineage and the Pacific robin radiation-a node that is poorly supported by mtDNA. We also find discordance in support for a sister relationship between the similarly plumaged Rose Robin (P. rosea) and Pink Robin (P. rodinogaster). Our nuclear data complement previous mtDNA studies in suggesting that the phenotypically cryptic eastern and western populations of Australia's Scarlet Robin (P. boodang) are genetically distinct lineages at the early stages of divergence and speciation.

Time-stepping techniques to enable the simulation of bursting behavior in a physiologically realistic computational islet.

Physiologically realistic simulations of computational islets of beta cells require the long-time solution of several thousands of coupled ordinary differential equations (ODEs), resulting from the combination of several ODEs in each cell and realistic numbers of several hundreds of cells in an islet. For a reliable and accurate solution of complex nonlinear models up to the desired final times on the scale of several bursting periods, an appropriate ODE solver designed for stiff problems is eventually a necessity, since other solvers may not be able to handle the problem or are exceedingly inefficient. But stiff solvers are potentially significantly harder to use, since their algorithms require at least an approximation of the Jacobian matrix. For sophisticated models, systems of several complex ODEs in each cell, it is practically unworkable to differentiate these intricate nonlinear systems analytically and to manually program the resulting Jacobian matrix in computer code. This paper demonstrates that automatic differentiation can be used to obtain code for the Jacobian directly from code for the ODE system, which allows a full accounting for the sophisticated model equations. This technique is also feasible in source-code languages Fortran and C, and the conclusions apply to a wide range of systems of coupled, nonlinear reaction equations. However, when we combine an appropriately supplied Jacobian with slightly modified memory management in the ODE solver, simulations on the realistic scale of one thousand cells in the islet become possible that are several orders of magnitude faster than the original solver in the software Matlab, a language that is particularly user friendly for programming complicated model equations. We use the efficient simulator to analyze electrical bursting and show non-monotonic average burst period between fast and slow cells for increasing coupling strengths. We also find that interestingly, the arrangement of the connected fast and slow heterogeneous cells impacts the peak bursting period monotonically.