Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Normal Distribution”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 505 records · Page 28
All sourcesNASA NTRSDOE OSTI
DOE OSTI2016

Jeans instability in collisional strongly coupled dusty plasma with radiative condensation and polarization force

The influence of dust-neutral collisions, polarization force, and electron radiative condensation is analysed on the Jeans (gravitational) instability of partially ionized strongly coupled dusty plasma (SCDP) using linear perturbation (normal mode) analysis. The Boltzmann distributed ions, dynamics of inertialess electrons, charged dust and neutral particles are considered. Using the plane wave solutions, a general dispersion relation is derived which is modified due to the presence of dust-neutral collisions, strong coupling effect, polarization force, electron radiative condensation, and Jeans dust/neutral frequencies. In the long wavelength perturbations, the Jeans instability criterion depends upon strong coupling effect, polarization interaction parameter, and thermal loss, but it is independent of dust-neutral collision frequency. The stability of the considered configuration is analysed using the Routh–Hurwitz criterion. The growth rates of Jeans instability are illustrated, and stabilizing influence of viscoelasticity and dust-neutral collision frequency while destabilizing effect of electron radiative condensation, polarization force, and Jeans dust-neutral frequency ratio is observed. This work is applied to understand the gravitational collapse of SCDP with dust-neutral collisions.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2022

Accelerating astronomical and cosmological inference with preconditioned Monte Carlo

ABSTRACT We introduce preconditioned Monte Carlo (PMC), a novel Monte Carlo method for Bayesian inference that facilitates efficient sampling of probability distributions with non-trivial geometry. PMC utilizes a Normalizing Flow (NF) in order to decorrelate the parameters of the distribution and then proceeds by sampling from the preconditioned target distribution using an adaptive Sequential Monte Carlo (SMC) scheme. The results produced by PMC include samples from the posterior distribution and an estimate of the model evidence that can be used for parameter inference and model comparison, respectively. The aforementioned framework has been thoroughly tested in a variety of challenging target distributions achieving state-of-the-art sampling performance. In the cases of primordial feature analysis and gravitational wave inference, PMC is approximately 50 and 25 times faster, respectively, than nested sampling (NS). We found that in higher dimensional applications, the acceleration is even greater. Finally, PMC is directly parallelisable, manifesting linear scaling up to thousands of CPUs.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2015

A statistical study of EMIC waves observed by Cluster. 1. Wave properties. EMIC Wave Properties

Electromagnetic ion cyclotron (EMIC) waves are an important mechanism for particle energization and losses inside the magnetosphere. In order to better understand the effects of these waves on particle dynamics, detailed information about the occurrence rate, wave power, ellipticity, normal angle, energy propagation angle distributions, and local plasma parameters are required. Previous statistical studies have used in situ observations to investigate the distribution of these parameters in the magnetic local time versus L-shell (MLT-L) frame within a limited magnetic latitude (MLAT) range. In our study, we present a statistical analysis of EMIC wave properties using 10 years (2001–2010) of data from Cluster, totaling 25,431 min of wave activity. Due to the polar orbit of Cluster, we are able to investigate EMIC waves at all MLATs and MLTs. This allows us to further investigate the MLAT dependence of various wave properties inside different MLT sectors and further explore the effects of Shabansky orbits on EMIC wave generation and propagation. Thus, the statistical analysis is presented in two papers. OUr paper focuses on the wave occurrence distribution as well as the distribution of wave properties. The companion paper focuses on local plasma parameters during wave observations as well as wave generation proxies.

58 GEOSCIENCES↗
DOE OSTI2025

Measurement of the top-quark pole mass in dileptonic $t\overline{t}$ + 1-jet events at $\sqrt{s}=13$ TeV with the ATLAS experiment

A measurement of the top-quark pole mass $m$$^{pole}_{t}$ is presented in $t\bar{t}$ events with an additional jet, $t\bar{t}$+ 1-jet, produced in pp collisions at $\sqrt{s} = 13 TeV. The data sample, recorded with the ATLAS experiment during Run 2 of the LHC, corresponds to an integrated luminosity of 140 fb −1 . Events with one electron and one muon of opposite electric charge in the final state are selected to measure the $t\bar{t}$ + 1-jet differential cross-section as a function of the inverse of the invariant mass of the $t\bar{t}$ + 1-jet system. Iterative Bayesian Unfolding is used to correct the data to enable comparison with fixed-order calculations at next-to-leading-order accuracy in the strong coupling. The process pp → $t\bar{t}$j(2 → 3), where top quarks are taken as stable particles, and the process pp → $b\bar{b}$l + vl – $\overline{ν}$j (2 → 7), which includes top-quark decays to the dilepton final state and off-shell effects, are considered. The top-quark mass is extracted using a χ 2 fit of the unfolded normalized differential cross-section distribution. The results obtained with the 2 → 3 and 2 → 7 calculations are compatible within theoretical uncertainties, providing an important consistency check.

Hadron-Hadron Scattering↗
DOE OSTI2015

Aerosol remote sensing in polar regions

Multi-year sets of ground-based sun-photometer measurements conducted at 12 Arctic sites and 9 Antarctic sites were examined to determine daily mean values of aerosol optical thickness τ(λ) at visible and near-infrared wavelengths, from which best-fit values of Ångström's exponent α were calculated. Analysing these data, the monthly mean values of τ(0.50 μm) and α and the relative frequency histograms of the daily mean values of both parameters were determined for winter–spring and summer–autumn in the Arctic and for austral summer in Antarctica. The Arctic and Antarctic covariance plots of the seasonal median values of α versus τ(0.50 μm) showed: (i) a considerable increase in τ(0.50 μm) for the Arctic aerosol from summer to winter–spring, without marked changes in α; and (ii) a marked increase in τ(0.50 μm) passing from the Antarctic Plateau to coastal sites, whereas α decreased considerably due to the larger fraction of sea-salt aerosol. Good agreement was found when comparing ground-based sun-photometer measurements of τ(λ) and α at Arctic and Antarctic coastal sites with Microtops measurements conducted during numerous AERONET/MAN cruises from 2006 to 2013 in three Arctic Ocean sectors and in coastal and off-shore regions of the Southern Atlantic, Pacific, and Indian Oceans, and the Antarctic Peninsula. Lidar measurements were also examined to characterise vertical profiles of the aerosol backscattering coefficient measured throughout the year at Ny-Ålesund. Satellite-based MODIS, MISR, and AATSR retrievals of τ(λ) over large parts of the oceanic polar regions during spring and summer were in close agreement with ship-borne and coastal ground-based sun-photometer measurements. An overview of the chemical composition of mode particles is also presented, based on in-situ measurements at Arctic and Antarctic sites. Fourteen log-normal aerosol number size-distributions were defined to represent the average features of nuclei, accumulation and coarse mode particles for Arctic haze, summer background aerosol, Asian dust and boreal forest fire smoke, and for various background austral summer aerosol types at coastal and high-altitude Antarctic sites. The main columnar aerosol optical characteristics were determined for all 14 particle modes, based on in-situ measurements of the scattering and absorption coefficients. Diurnally averaged direct aerosol-induced radiative forcing and efficiency were calculated for a set of multimodal aerosol extinction models, using various Bidirectional Reflectance Distribution Function models over vegetation-covered, oceanic and snow-covered surfaces. These gave a reliable measure of the pronounced effects of aerosols on the radiation balance of the surface–atmosphere system over polar regions.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2020

Network reconfiguration and distributed energy resource scheduling for improved distribution system resilience

Electric utility companies work to restore as much load as possible after power outages caused by extreme weather events. In this paper, an outage management strategy is proposed to enhance distribution system resilience through network reconfiguration and distributed energy resources (DERs) scheduling. After a line fault, the proposed algorithm can identify radial network topology based on the rank of the incidence matrix. The reconfiguration is implemented by switching tie lines and sectionalizing lines. With the new network topology, an optimal DER scheduling problem is solved to minimize the accumulative cost for dispatchable DER operation and load reduction. Finally, the optimal topology that minimizes the accumulative cost is selected from all radial topologies. The computational workload is relatively low because only linear programming needs to be solved. Using the case studies of the IEEE 69-bus and IEEE 123-bus systems, we consider the worst-case scenarios in which faults occur in the upstream feeder. The simulation results demonstrate that the proposed strategy allows for a relatively high percentage of the load to remain in service after line faults. Furthermore, compared with microgrid-formation approaches, the proposed strategy has advantages when applied to the distribution systems with several normally-open tie lines and low DER penetration.

42 ENGINEERING↗
DOE OSTI2024

A new portable random number generator wrapper library

Random number generator is an important component of many scientific projects. Many projects are written using programming models (like OpenMP and SYCL) to target different architectures. However, some programming models do not provide a random number generator. In this work, we introduce our random number generator wrapper. It is a header-only library that supports three distributions of random numbers: uniform, normal, and poisson. On the GPU backend, it wraps the cuRAND and rocRAND library, and supports various random number engines. It also wraps random123, a counterbased random number generator, on both CPU and GPU. With this library, we can generate random numbers with a few lines of code and target both GPU and multi-thread CPU with the same code. We also investigate the performance and scalability of this wrapper on different architectures with different engines and the number of cores.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2014

Non-local thermodynamic equilibrium 1.5D modeling of red giant stars

Spectra for two-dimensional (2D) stars in the 1.5D approximation are created from synthetic spectra of one-dimensional (1D) non-local thermodynamic equilibrium (NLTE) spherical model atmospheres produced by the PHOENIX code. The 1.5D stars have the spatially averaged Rayleigh-Jeans flux of a K3-4 III star while varying the temperature difference between the two 1D component models (ΔT {sub 1.5D}) and the relative surface area covered. Synthetic observable quantities from the 1.5D stars are fitted with quantities from NLTE and local thermodynamic equilibrium (LTE) 1D models to assess the errors in inferred T {sub eff} values from assuming horizontal homogeneity and LTE. Five different quantities are fit to determine the T {sub eff} of the 1.5D stars: UBVRI photometric colors, absolute surface flux spectral energy distributions (SEDs), relative SEDs, continuum normalized spectra, and TiO band profiles. In all cases except the TiO band profiles, the inferred T {sub eff} value increases with increasing ΔT {sub 1.5D}. In all cases, the inferred T {sub eff} value from fitting 1D LTE quantities is higher than from fitting 1D NLTE quantities and is approximately constant as a function of ΔT {sub 1.5D} within each case. The difference between LTE and NLTE for the TiO bands is caused indirectly by the NLTE temperature structure of the upper atmosphere, as the bands are computed in LTE. We conclude that the difference between T {sub eff} values derived from NLTE and LTE modeling is relatively insensitive to the degree of the horizontal inhomogeneity of the star being modeled and largely depends on the observable quantity being fit.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2017

Light-front spin-dependent spectral function and nucleon momentum distributions for a three-body system

Poincare covariant definitions for the spin-dependent spectral function and for the momentum distributions within the light-front Hamiltonian dynamics are proposed for a three-fermion bound system, starting from the light-front wave function of the system. The adopted approach is based on the Bakamjian–Thomas construction of the Poincaré generators, which allows one to easily import the familiar and wide knowledge on the nuclear interaction into a light-front framework. The proposed formalism can find useful applications in refined nuclear calculations, such as those needed for evaluating the European Muon Collaboration effect or the semi-inclusive deep inelastic cross sections with polarized nuclear targets, since remarkably the light-front unpolarized momentum distribution by definition fulfills both normalization and momentum sum rules. As a result, also shown is a straightforward generalization of the definition of the light-front spectral function to an A -nucleon system.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2020

Measurements of $\pi^-$ production in $^7$Be+$^9$Be collisions at beam momenta from 19$A$ to 150$A$GeV/$c$ in the NA61/SHINE experiment at the CERN SPS

The NA61/SHINE collaboration studies at the CERN Super Proton Synchrotron (SPS) the onset of deconfinement in hadronic matter by the measurement of particle production in collisions of nuclei with various sizes at a set of energies covering the SPS energy range. This paper presents results on inclusive double-differential spectra and mean multiplicities of $\pi^{-}$ mesons produced in the 5% most central $^7$Be + $^9$Be collisions at beam momenta of 19$A$, 30$A$, 40$A$, 75$A$ and 150$A$ GeV/$c$ obtained by the so-called $h^-$ method which does not require any particle identification. The shape of the transverse mass spectra differs from the shapes measured in central Pb + Pb collisions and inelastic p + p interactions. The normalized width of the rapidity distribution decreases with increasing collision energy and is in between the results for inelastic nucleon-nucleon and central Pb + Pb collisions. The mean multiplicity of pions per wounded nucleon in central $^7$Be + $^9$Be collisions is close to that in central Pb + Pb collisions up to 75$A$GeV/$c$. However, at the top SPS energy the result is close to the one for inelastic nucleon-nucleon interactions. The results are discussed in the context of predictions for the onset of deconfinement at the CERN SPS collision energies.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2017

Multilevel Sequential Monte Carlo Samplers for Normalizing Constants

This article considers the sequential Monte Carlo (SMC) approximation of ratios of normalizing constants associated to posterior distributions which in principle rely on continuum models. Therefore, the Monte Carlo estimation error and the discrete approximation error must be balanced. A multilevel strategy is utilized to substantially reduce the cost to obtain a given error level in the approximation as compared to standard estimators. Two estimators are considered and relative variance bounds are given. The theoretical results are numerically illustrated for two Bayesian inverse problems arising from elliptic partial differential equations (PDEs). The examples involve the inversion of observations of the solution of (i) a 1-dimensional Poisson equation to infer the diffusion coefficient, and (ii) a 2-dimensional Poisson equation to infer the external forcing.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

QuantumNAT: quantum noise-aware training with noise injection, quantization and normalization

Parameterized Quantum Circuits (PQC) are promising towards quantum advantage on near-term quantum hardware. However, due to the large quantum noises (errors), the performance of PQC models has a severe degradation on real quantum devices. Take Quantum Neural Network (QNN) as an example, the accuracy gap between noise-free simulation and noisy results on IBMQ-Yorktown for MNIST-4 classification is over 60%. Existing noise mitigation methods are general ones without leveraging unique characteristics of PQC; on the other hand, existing PQC work does not consider noise effect. To this end, we present QuantumNAT, a PQC-specific framework to perform noise-aware optimizations in both training and inference stages to improve robustness. We experimentally observe that the effect of quantum noise to PQC measurement outcome is a linear map from noise-free outcome with a scaling and a shift factor. Motivated by that, we propose post-measurement normalization to mitigate the feature distribution differences between noise-free and noisy scenarios. Furthermore, to improve the robustness against noise, we propose noise injection to the training process by inserting quantum error gates to PQC according to realistic noise models of quantum hardware. Lastly, post-measurement quantization is introduced to quantize the measurement outcomes to discrete values, achieving the denoising effect. Extensive experiments on 8 classification tasks using 6 quantum devices demonstrate that QuantumNAT improves accuracy by up to 43%, and achieves over 94% 2-class, 80% 4-class, and 34% 10-class classification accuracy measured on real quantum computers. The code for construction and noise-aware training of PQC is available in the TorchQuantum library.

Wang, Hanrui↗
DOE OSTI2017

Correction of confidence intervals in excess relative risk models using Monte Carlo dosimetry systems with shared errors

Introduction: Measurement errors are ubiquitous in epidemiological studies, especially for doses arising from environmental and occupational exposures, which are difficult to assess. Correction methods for independent measurement errors, have been comprehensively described elsewhere, including regression calibration, simulation extrapolation (SIMEX), etc. However, when exposure estimates are constructed using complicated physical and biological models, uncertainties in the dosimetry system can become very complex. In such systems, uncertainties in some parameters may affect a large group of study participants simultaneously, and hence are “shared” as opposed to independent dosimetric uncertainties. Representing the structure of the uncertainty in the dosimetry by providing multiple realizations of possible dose has been described for several studies. The goal of this paper is to implement, and explore the properties of, a generalized estimating equation-based approach to shared uncertainties based on principles described in an earlier paper. Methods: The effects of shared and unshared dosimetry error on parameter estimates and confidence intervals in a generalized excess relative risk model are examined. A post-hoc correction method is implemented that constructs corrected confidence intervals based on an underlying variance structure described in an earlier paper. In this correction method it is assumed that the estimated slope parameter in a linear excess relative risk model is distributed as a mixture of normal and lognormal components. A simulation study of the behavior of this correction was developed based on multiple realizations provided by the Monte Carlo dosimetry system used for the plutonium exposures in the Mayak Worker Cohort and a risk model with close similarity to that used in analyses of lung cancer data in that cohort. Results: In the simulation study we found that shared dose errors severely degraded the coverage probabilities of confidence limits for the slope parameter in the linear excess relative risk model relating uncertain exposure to lung cancer risk. Using our correction method to adjust the confidence intervals provided much improved coverage probabilities for this parameter. A smaller amount of degradation in confidence interval coverage for other parameters (e.g. those in the baseline and dose modification terms) was noted, and the proposed correction method also improved coverage for these parameters as well. Discussion: Compared to other recently described methods our proposed correction method relies on relatively familiar software to fit dose response models. The basic modeling philosophy described here first uses existing Poisson regression methods to estimate parameters in the excess relative risk model; in this first stage mean dose (computed from the multiple realizations) is used as the independent variable in the regression. Once the model is fit we then correct (widen) the standard errors for all parameters using the proposed correction method. The simulation studies indicate that the methods described here are an important contribution to the available tools for dealing with shared measurement errors in complex risk analysis.

61 RADIATION PROTECTION AND DOSIMETRY↗
DOE OSTI2024

Relativistic Alfvén Turbulence at Kinetic Scales

Abstract In a strongly magnetized, magnetically dominated relativistic plasma, Alfvénic turbulence can extend to scales much smaller than the particle inertial scales. It leads to an energy cascade somewhat analogous to inertial- or kinetic-Alfvén turbulent cascades existing in nonrelativistic space and astrophysical plasmas. Based on phenomenological modeling and particle-in-cell numerical simulations, we propose that the energy spectrum of such relativistic kinetic-scale Alfvénic turbulence is close to k −3 or slightly steeper than that due to intermittency corrections or Landau damping. We note the analogy of this spectrum with the Kraichnan spectrum corresponding to the enstrophy cascade in 2D incompressible fluid turbulence. Such turbulence strongly energizes particles in the direction parallel to the background magnetic field, leading to nearly one-dimensional particle momentum distributions. We find that these distributions have universal log-normal statistics.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2016

Chattanooga Electric Power Board Case Study Distribution Automation

In 2009, the U.S. Department of Energy under the American Recovery and Reinvestment Act (ARRA) awarded a grant to the Chattanooga, Tennessee, Electric Power Board (EPB) as part of the Smart Grid Investment Grant Program. The grant had the objective “to accelerate the transformation of the nation’s electric grid by deploying smart grid technologies.” This funding award enabled EPB to expedite the original smart grid implementation schedule from an estimated 10-12 years to 2.5 years. With this funding, EPB invested heavily in distribution automation technologies including installing over 1,200 automated circuit switches and sensors on 171 circuits. For utilities considering a commitment to distribution automation, there are underlying questions such as the following: “What is the value?” and “What are the costs?” This case study attempts to answer these questions. The primary benefit of distribution automation is increased reliability or reduced power outage duration and frequency. Power outages directly impact customer economics by interfering with business functions. In the past, this economic driver has been difficult to effectively evaluate. However, as this case study demonstrates, tools and analysis techniques are now available. In this case study, the impact on customer costs associated with power outages before and after the implementation of distribution automation are compared. Two example evaluations are performed to demonstrate the benefits: 1) a savings baseline for customers under normal operations1 and 2) customer savings for a single severe weather event. Cost calculations for customer power outages are performed using the US Department of Energy (DOE) Interruption Cost Estimate (ICE) calculator2. This tool uses standard metrics associated with outages and the customers to calculate cost impact. The analysis shows that EPB customers have seen significant reliability improvements from the implementation of distribution automation. Under normal operations, the investment in distribution automation has enabled a 43.5% reduction in annual outage minutes since 2012. This has led to an estimated total savings of $26.8 million per year. Examining a single severe weather event3, the distribution automation was able to restore power to 40,579 (nearly 56%) customers within 1–2 seconds and reduce outage minutes by 29.0%. This saved customers an estimated $23.2 million over the course of the storm.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2016

SU-F-T-565: Assessment of Dosimetric Accuracy for a 3D Gel-Based Dosimetry Service

Purpose: To assess the 3D dosimetric accuracy when using a mail-in service for square and stereotactic fields in a clinical environment. Methods: The 3D dosimetry mail-in service (3DDaaS), offered by Modus QA (London, ON), was used to measure dose distributions from a 6 MV beam of a Varian Clinac. Plastic jars filled with radiosensitive ClearView™ gel were received, CT scanned (for registration and density information), irradiated, and then mailed back to the manufacturer for optical CT readout. Three square field irradiations (2×2, 4×4, and 10×10 cm{sup 2}) were performed with jars immobilized in a water tank, and a composite small-field stereotactic delivery was performed using an in-air holder. Dosimetric properties of the gel were quantified within the 25–50 Gy dose range using 3D optical attenuation (OA) distributions provided by the manufacturer. OA was normalized to 100% at the position of isocenter, which received 40Gy. Percentage depth dose, profiles, and 3D gamma distributions (3%/1mm criteria) were calculated to quantify feasibility for relative dosimetry. Results: Mean CT-measured density in the central (3×3×3) cm{sup 3} gel region was 40 ± 3 HU, indicating good homogeneity and near-water-equivalence. Measured and calculated central axis doses agreed to within ±3% in the 25–50 Gy dose range. For the square field irradiations, dose profiles agreed to within 1mm. Gamma analysis of the composite irradiation yielded 99.8%, 91.4%, and 79.1% passing rates for regions receiving at least 10, 5, and 2 Gy, respectively, indicating feasibility for use in high-dose regions. Absolute response varied by up to 16% between jars, indicating limitations for absolute dosimetry under the mail-in conditions. Conclusion: 3DDaaS is a novel near-water-equivalent dosimetry system accurate to within 3% dose and 1mm 3D spatial resolution, and is straightforward to use in a clinical setting. Future investigations are warranted to improve dosimeter response in low-dose regions. The authors would like to thank ModusQA (London, ON) for providing the gels and optical readouts used this work. This work was partially funded by NIH P01CA059827.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2022

Microhardness Distribution of Long Magnesium Block Processed through Powder Metallurgy

Powder metallurgy is a popular method of making raw powders into specific shaped samples. However, the pressure distribution and the microhardness difference within the sample are nonnegligible and unclear when the sample is long or exceeds a specific size. In this study, the long magnesium blocks, with a ratio of about 2.8 between the sample height and the sample side length, are successfully synthesized under three uniaxial and two biaxial conditions. Then, the sample hardness values on the outer surface and the center plane are tested to study the microhardness distribution. The modified analytical expression indicates that the normal pressure exponentially decreases along the compression direction, which is consistent with the hardness distribution trend. Because higher pressure leads to a more compact arrangement of the powders, more metal bonds are formed after sintering. During the first pressing, the sidewall pressure makes the surface hardness higher. The secondary reverse compression mainly improves the bottom and core hardness due to the re-orientation and re-location of the powders. The obtained relationship between the applied pressure and the hardness distribution is instructive in predicting and improving the sample quality.

Wang, Jiaying↗
DOE OSTI2016

TH-CD-209-09: Quickly Identifying Good Candidates for Proton Therapy From Geometric Considerations

Purpose: We developed a knowledge-based model that can predict the patient-specific benefits of proton therapy based upon geometric considerations. The model could also aid patient selection in model-based clinical trials or help justify clinical decisions to insurance companies. Methods: The knowledge-based method trains a model upon existing proton treatment plans, exploiting correlations between dose and distance-to-target. Each OAR is split into concentric subvolumes surrounding the target volume, and a skew-normal PDF is fit to the dose distribution found within each shell. The model learns from shared trends in how the best-fit skew-normal parameters depend upon distance-to-target. It can then predict feasible OAR DVHs for a new patient (without a proton plan) based upon their geometry. The expected benefits of proton therapy are assessed by comparing the predicted DVHs to those of an IMRT plan, using a metric such as the equivalent uniform dose (EUD). Results: A model was trained for clival chordoma, owing to its geometric complexity and the multitude of nearby OARs. The model was trained using 20 patients and validated with a further 20 patients, and considers several different OARs. The predicted EUD was in good agreement with that of the actual proton plan. The coefficient of determination (R-squared) was 85% overall, 92% for cochleas, 80% for optic chiasm and 79% for spinal cord. The model exhibited no signs of bias or overfitting. When compared to an IMRT plan, the model could classify whether a patient will experience a gain or a loss with an accuracy between 75% and 95%, depending upon the OAR. Conclusion: We developed a model that can quickly and accurately predict the patient-specific benefits of proton therapy in clival chordoma patients, though models could be trained for other tumor sites. This work is funded by National Cancer Institute grant U19 CA 021239.

60 APPLIED LIFE SCIENCES↗