Search NASA⌕ Search

SEARCH · Search NASA

Results for “Marginal 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 289 records · Page 16

Adversarial classification via distributional robustness with Wasserstein ambiguity

Abstract We study a model for adversarial classification based on distributionally robust chance constraints. We show that under Wasserstein ambiguity, the model aims to minimize the conditional value-at-risk of the distance to misclassification, and we explore links to adversarial classification models proposed earlier and to maximum-margin classifiers. We also provide a reformulation of the distributionally robust model for linear classification, and show it is equivalent to minimizing a regularized ramp loss objective. Numerical experiments show that, despite the nonconvexity of this formulation, standard descent methods appear to converge to the global minimizer for this problem. Inspired by this observation, we show that, for a certain class of distributions, the only stationary point of the regularized ramp loss minimization problem is the global minimizer.

Ho-Nguyen, Nam (ORCID:0000000344647730)↗

Power Distribution Control Strategy of a Conceptual Soluble Boron Free SMR

In a commercial PWR, the two primary functions of control rods are to provide rapid shutdown and to control power distribution, whereas the initial excess reactivity associated with fuel loading is controlled by the soluble boron in the coolant and burnable absorbers. The required number of total control rods is determined by the shutdown margin of the reactor. The control rod grouping is also required to control power distribution. Therefore, removal of soluble boron requires extra reliance on the control rods to compensate excess reactivity. In a soluble boron free reactor which has highly negative MTC requires an increase in both worth and number of control rod assemblies to ensure adequate shutdown margin. In addition, since control rods are partially inserted into the reactor core, their presence adversely affects the power distribution and thus may make high peaking factors leading to an operating margin reduction. In this paper, the strategy of control rods operation to control excess reactivity and power distribution have been studied. Also, shutdown capability of the control rods was examined. A loading pattern for a Conceptual Soluble Boron Free Small Modular Reactor (CSBF-SMR) has been constructed. Two groups of regulating banks employ Ag-In-Cd to control excess reactivity and power distributions during the cycle operation and 1 group of shutdown bank made of B{sub 4}C were used.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Aeolian Sediment Transport Pathways and Aerodynamics at Troughs on Mars

Interaction between wind regimes and topography can give rise to complex suites of aeolian landforms. This paper considers aeolian sediment associated wit11 troughs on Mars and identifies a wider range of deposit types than has previously been documented. These include wind streaks, falling dunes, "lateral" dunes, barchan dunes, linear dunes, transverse ridges, sand ramps, climbing dunes, sand streamers, and sand patches. The sediment incorporated into these deposits is supplied by wind streaks and ambient Planitia sources as well as originating within the trough itself, notably from the trough walls and floor. There is also transmission of sediment between d~~neTsh. e flow dynamics which account for the distribution of aeolian sediment have been modeled using two-dimensional computational fluid dynamics. The model predicts flow separation on the upwind side of the trough followed by reattachment and acceleration at the downwind margin. The inferred patterns of sediment transport compare well with the distribution of aeolian forms. Model data indicate an increase of wind velocity by approx. 30 % at the downwind trough margin. This suggests that the threshold wind speed necessary for sand mobilization on Mars will be more freq~~entmlye t in these inclined locations.

Bourke, Mary C.↗

Stimulated emission of AKR in regions of a weakly unstable electron velocity distribution

While, according to the cyclotron maser theory, auroral kilometric radiation generates auroral zone instability whose consequent radiation is in many respects consistent with observations, in situ velocity distribution measurements by satellites show only moderate positive gradients which would be only marginally unstable, implying a low radiation level. Numerical simulations are conducted alternatively using an observed electron-velocity distribution and an idealized loss-cone distribution. The expected amplification of X-mode radiation, as well as effects which go beyond the behavior of a simple amplifying medium, emerge in the results.

Kainer, S.↗

A Framework for Assessing Economic and Environmental Trade-offs of Internalized Emission Costs in ERCOT Grid Planning

The power grid is on the cusp of a massive transition driven by three major areas: 1) the growth in demand for electricity, 2) efforts to decarbonize the United States economy, and 3) a desire to mitigate social disparities from the impact of electricity generation on local populations. However, most studies of the electricity sector do not include equity impacts in their models. This study seeks to do so by developing a comprehensive and generalizable model tailored to the Electric Reliability Council of Texas (ERCOT) grid, designed to incorporate the equity impacts of electricity generation in a decarbonized and resilient framework. To integrate equity into our research, we incorporate environmental externalities into our capacity expansion model of ERCOT. Specifically, we factor in intermediate-level local marginal damages of precursor pollutants (NH3, NOx, primary PM2.5, SO2, and VOC) and global pollutant CO2 into the cost of generating electricity. We do this by taking into account county population, county ambient pollution concentration, and generator emission rates. Leveraging open-source modeling tools, such as PowerGenome, pyGRETA, and GenX we construct a county-level model to account for these costs. We integrate these marginal damages into the variable operations and maintenance costs of generators, for both existing and potential future builds. This study’s findings suggest that the value of a dynamic social cost of carbon (SSC) will cover criteria pollutant marginal damages within the ERCOT grid and solar and wind is expected to increase out to 2035. Key metrics evaluated within this research include fuel mix distribution across technologies, transmission and grid infrastructure costs, CO2 emissions, local pollutants marginal damages, and the variation in generation capacity built by the model. These results and framework can be used to support grid decisions that explicitly include distributional and procedural equity within a decarbonized and sustainable grid framework.

24 POWER TRANSMISSION AND DISTRIBUTION↗

JET PROPERTIES OF GeV-SELECTED RADIO-LOUD NARROW-LINE SEYFERT 1 GALAXIES AND POSSIBLE CONNECTION TO THEIR DISK AND CORONA

The observed spectral energy distributions of five GeV-selected narrow-line Seyfert 1 (NLS1) galaxies are fitted with a model including the radiation ingredients from the relativistic jet, the accretion disk, and the corona. We compare the properties of these GeV NLS1 galaxies with flat spectrum radio quasars (FSRQs), BL Lacertae objects (BL Lacs), and radio-quiet (RQ) Seyfert galaxies, and explore possible hints for jet-disk/corona connection. Our results show that the radiation physics and the jet properties of the GeV NLS1 galaxies resemble that of FSRQs. The luminosity variations of PMN J0948+0022 and 1H 0323+342 at the GeV band is tightly correlated with the beaming factor (δ), similar to that observed in FSRQ 3C 279. The accretion disk luminosities and the jet powers of the GeV NLS1 galaxies cover both the ranges of FSRQs and BL Lacs. With the detection of bright corona emission in 1H 0323+342, we show that the ratio of the corona luminosity (L {sub corona}) to the accretion disk luminosity (L {sub d}) is marginally within the high end of this ratio distribution for an RQ Seyfert galaxy sample, and the variation of jet luminosity may connect with L {sub corona}. However, it is still unclear whether a system with a high L {sub corona}/L {sub d} ratio prefers to power a jet.

79 ASTRONOMY AND ASTROPHYSICS↗

Radiotherapy margin design with particular consideration of high curvature CTVs

In applying 3D conformal radiation therapy to a tumor clinical target volume (CTV), a margin is added around the CTV to account for any sources of error in the application of treatment which may result in misalignment between the CTV and the dose distribution actually delivered. The volume enclosed within the CTV plus the margin is known as the PTV, or planning target volume. The larger the errors are anticipated to be, the wider the margin will need to be to accommodate those errors. Based on the approach of van Herk et al. [''The probability of correct target dosage: Dose-population histograms for deriving treatment margins in radiotherapy,'' Int. J. Radiat. Oncol. Biol., Phys. 47(4), 1121-1135 (2000)] this paper develops the mathematical theory behind the calculation of the margin width required to ensure that the entire CTV receives sufficiently high dose with sufficiently high probability. The margin recipe developed not only considers the magnitude of the errors but also includes a term to adjust for curved CTV surfaces. In doing so, the accuracy of the margin recipe is enhanced yet remains mathematically concise enough to be readily implemented in the clinical setting. The results are particularly relevant for clinical situations in which the uncertainties in treatment are large relative to the size of the CTV.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

EFFECTS OF DYNAMICAL EVOLUTION OF GIANT PLANETS ON SURVIVAL OF TERRESTRIAL PLANETS

The orbital distributions of currently observed extrasolar giant planets allow marginally stable orbits for hypothetical, terrestrial planets. In this paper, we propose that many of these systems may not have additional planets on these ''stable'' orbits, since past dynamical instability among giant planets could have removed them. We numerically investigate the effects of early evolution of multiple giant planets on the orbital stability of the inner, sub-Neptune-like planets which are modeled as test particles, and determine their dynamically unstable region. Previous studies have shown that the majority of such test particles are ejected out of the system as a result of close encounters with giant planets. Here, we show that secular perturbations from giant planets can remove test particles at least down to 10 times smaller than their minimum pericenter distance. Our results indicate that, unless the dynamical instability among giant planets is either absent or quiet like planet-planet collisions, most test particles down to {approx}0.1 AU within the orbits of giant planets at a few AU may be gone. In fact, out of {approx}30% of survived test particles, about three quarters belong to the planet-planet collision cases. We find a good agreement between our numerical results and the secular theory, and present a semi-analytical formula which estimates the dynamically unstable region of the test particles just from the evolution of giant planets. Finally, our numerical results agree well with the observations, and also predict the existence of hot rocky planets in eccentric giant planet systems.

79 ASTRONOMY AND ASTROPHYSICS↗

Predictions from star formation in the multiverse

We compute trivariate probability distributions in the landscape, scanning simultaneously over the cosmological constant, the primordial density contrast, and spatial curvature. We consider two different measures for regulating the divergences of eternal inflation, and three different models for observers. In one model, observers are assumed to arise in proportion to the entropy produced by stars; in the others, they arise at a fixed time (5 or 10x10{sup 9} years) after star formation. The star formation rate, which underlies all our observer models, depends sensitively on the three scanning parameters. We employ a recently developed model of star formation in the multiverse, a considerable refinement over previous treatments of the astrophysical and cosmological properties of different pocket universes. For each combination of observer model and measure, we display all single and bivariate probability distributions, both with the remaining parameter(s) held fixed and marginalized. Our results depend only weakly on the observer model but more strongly on the measure. Using the causal diamond measure, the observed parameter values (or bounds) lie within the central 2{sigma} of nearly all probability distributions we compute, and always within 3{sigma}. This success is encouraging and rather nontrivial, considering the large size and dimension of the parameter space. The causal patch measure gives similar results as long as curvature is negligible. If curvature dominates, the causal patch leads to a novel runaway: it prefers a negative value of the cosmological constant, with the smallest magnitude available in the landscape.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Quantification of the Variability of Diaphragm Motion and Implications for Treatment Margin Construction

Purpose: To quantify the variability of diaphragm motion during free-breathing radiotherapy of lung patients and its effect on treatment margins to account for geometric uncertainties. Methods and Materials: Thirty-three lung cancer patients were analyzed. Each patient had 5-19 cone-beam scans acquired during different treatment fractions. The craniocaudal position of the diaphragm dome on the same side as the tumor was tracked over 2 min in the projection images, because it is both easily visible and a suitable surrogate to study the variability of the tumor motion and its impact on treatment margins. Intra-acquisition, inter-acquisition, and inter-patient variability of the respiratory cycles were quantified separately, as were the probability density functions (PDFs) of the diaphragm position over each cycle, each acquisition, and each patient. Asymmetric margins were simulated using each patient PDF and compared to symmetric margins computed from a margin recipe. Results: The peak-to-peak amplitude variability (1 SD) was 3.3 mm, 2.4 mm, and 6.1 mm for the intra-acquisition, inter-acquisition, and inter-patient variability, respectively. The average PDF of each cycle was similar to the sin{sup 4} function but the PDF of each acquisition was closer to a skew-normal distribution because of the motion variability. Despite large interfraction baseline variability, the PDF of each patient was generally asymmetric with a longer end-inhale tail because the end-exhale position was more stable than the end-inhale position. The asymmetry of the PDF required asymmetric margins around the time-averaged position to account for the position uncertainty but the average difference was 1.0 mm (range, 0.0-4.4 mm) for a sharp penumbra and an idealized online setup correction protocol. Conclusion: The respiratory motion is more irregular during the fractions than between the fractions. The PDF of the respiratory motion is asymmetrically distributed. Both the intra-acquisition variability and the PDF asymmetry have a limited impact on dose distributions and inferred margins. The use of a margin recipe to account for respiratory motion with an estimate of the average motion amplitude was adequate in almost all patients.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Practical Method of Adaptive Radiotherapy for Prostate Cancer Using Real-Time Electromagnetic Tracking

Purpose: We have created an automated process using real-time tracking data to evaluate the adequacy of planning target volume (PTV) margins in prostate cancer, allowing a process of adaptive radiotherapy with minimal physician workload. We present an analysis of PTV adequacy and a proposed adaptive process. Methods and Materials: Tracking data were analyzed for 15 patients who underwent step-and-shoot multi-leaf collimation (SMLC) intensity-modulated radiation therapy (IMRT) with uniform 5-mm PTV margins for prostate cancer using the Calypso Registered-Sign Localization System. Additional plans were generated with 0- and 3-mm margins. A custom software application using the planned dose distribution and structure location from computed tomography (CT) simulation was developed to evaluate the dosimetric impact to the target due to motion. The dose delivered to the prostate was calculated for the initial three, five, and 10 fractions, and for the entire treatment. Treatment was accepted as adequate if the minimum delivered prostate dose (D{sub min}) was at least 98% of the planned D{sub min}. Results: For 0-, 3-, and 5-mm PTV margins, adequate treatment was obtained in 3 of 15, 12 of 15, and 15 of 15 patients, and the delivered D{sub min} ranged from 78% to 99%, 96% to 100%, and 99% to 100% of the planned D{sub min}. Changes in D{sub min} did not correlate with magnitude of prostate motion. Treatment adequacy during the first 10 fractions predicted sufficient dose delivery for the entire treatment for all patients and margins. Conclusions: Our adaptive process successfully used real-time tracking data to predict the need for PTV modifications, without the added burden of physician contouring and image analysis. Our methods are applicable to other uses of real-time tracking, including hypofractionated treatment.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Characterization of Partially Observed Epidemics - Application to COVID-19

This report documents a statistical method for the "real-time" characterization of partially observed epidemics. Observations consist of daily counts of symptomatic patients, diagnosed with the disease. Characterization, in this context, refers to estimation of epidemiological parameters that can be used to provide short-term forecasts of the ongoing epidemic, as well as to provide gross information for the time-dependent infection rate. The characterization problem is formulated as a Bayesian inverse problem, and is predicated on a model for the distribution of the incubation period. The model parameters are estimated as distributions using a Markov Chain Monte Carlo (MCMC) method, thus quantifying the uncertainty in the estimates. The method is applied to the COVID-19 pandemic of 2020, using data at the country, provincial (e.g., states) and regional (e.g. county) levels. The epidemiological model includes a stochastic component due to uncertainties in the incubation period. This model-form uncertainty is accommodated by a pseudo-marginal Metropolis-Hastings MCMC sampler, which produces posterior distributions that reflect this uncertainty. We approximate the discrepancy between the data and the epidemiological model using Gaussian and negative binomial error models; the latter was motivated by the over-dispersed count data. For small daily counts we find the performance of the calibrated models to be similar for the two error models. For large daily counts the negative-binomial approximation is numerically unstable unlike the Gaussian error model. Application of the model at the country level (for the United States, Germany, Italy, etc.) generally provided accurate forecasts, as the data consisted of large counts which suppressed the day-to-day variations in the observations. Further, the bulk of the data is sourced over the duration before the relaxation of the curbs on population mixing, and is not confounded by any discernible country-wide second wave of infections. At the state-level, where reporting was poor or which evinced few infections (e.g., New Mexico), the variance in the data posed some, though not insurmountable, difficulties, and forecasts were able to capture the data with large uncertainty bounds. The method was found to be sufficiently sensitive to discern the flattening of the infection and epidemic curve due to shelter-in-place orders after around 90% quantile for the incubation distribution (about 10 days for COVID-19). The proposed model was also used at a regional level to compare the forecasts for the central and north-west regions of New Mexico. Modeling the data for these regions illustrated different disease spread dynamics captured by the model. While in the central region the daily counts peaked in the late April, in the north-west region the ramp-up continued for approximately three more weeks.

59 BASIC BIOLOGICAL SCIENCES↗

Gains From Real-Time Tracking of Prostate Motion During External Beam Radiation Therapy

Purpose: To study the gains from real-time tracking of prostate motion and threshold-based intervention and the feasibility of margin reduction for external beam radiation therapy of prostate cancer. Methods and Materials: Prostate intrafractional motion data from 775 randomly selected treatment fractions (105 prostate patients) were analyzed. Statistical distributions of prostate intrafractional displacement from baseline were used for treatment margin calculation together with other geometrical uncertainties for all patients and a subset of 7 patient who exhibited the largest intrafractional motion. Compared with treatment without any intrafractional intervention, potential reductions in treatment margins were evaluated for treatments with 5-mm and 3-mm threshold-based intervention and four-dimensional (4D) treatments with and without prostate rotation correction. Results: The percentage of time of prostate displacement from the baseline by 3 mm and 5 mm in any direction was 13.4% and 1.8%, respectively, for the general patient population. The ratios were 41% and 15% for the 7 selected patients. Reductions in the posterior margin were 0.2, 0.5, 1.3, and 3.1 mm from the original 7.7 mm, respectively, for 5-mm and 3-mm threshold-based treatments and 4D treatments with and without prostate rotation correction for all patients. They were 1.3, 1.9, 3.1 and 4.9 mm from the original 9.5 mm, corresponding to the 7 selected patients. The treatment margin reductions in other directions were even smaller. Conclusions: Real-time motion tracking and threshold-based intrafractional intervention may play a significant roll in treatment margin reduction for a small fraction of patients but not for the general patient population. Four-dimensional treatments with prostate rotation correction can reduce the treatment margin more significantly.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Imaging Zn and Ni distributions in leaves of different ages of the hyperaccumulator Noccaea caerulescens by synchrotron-based X-ray fluorescence

Mapping of leaves of hyperaccumulators can provide insights into the mechanisms these species utilize to accumulate high metal concentrations. We used synchrotron-based X-ray fluorescence (SXRF) to perform Zn and Ni imaging in leaves of different ages of Noccaea caerulescens. A mature leaf of the related non-hyperaccumulator Thlaspi arvense was also imaged. The concentrations of Zn, Ni, Co, and Cr in N. caerulescens grown on an ultramafic soil were 9-, 10-, 12-, and 3-fold higher than T. arvense. N. caerulescens showed an exceptional ability to accumulate Zn from the soil, posing a bioconcentration factor of 6.7. T. arvense had Zn and Ni distributed uniformly in the leaf blade with doubling fluorescence counts in the tip and margins, suggesting a strategy to excrete metals and avoid toxicity. On the other hand, N. caerulescens displayed distinctly different Zn and Ni accumulation patterns, regardless of the age or metal concentration in the leaves. Zinc was mainly distributed in the cells surrounding the central and secondary veins. Nickel accumulated in the margins and tips of the leaf blade. Finally, given the time required to image large leaves in synchrotron facilities, small leaves can be used to represent the leaf distribution of Zn and Ni in N. caerulescens.

36 MATERIALS SCIENCE↗

Prediction of grain structure after thermomechanical processing of U-10Mo alloy using sensitivity analysis and machine learning surrogate model

Abstract Hot rolling and annealing are critical intermediate steps for controlling microstructures and thickness variations when fabricating uranium alloyed with 10% molybdenum (U-10Mo), which is highly relevant to worldwide nuclear non-proliferation efforts. This work proposes a machine-learning surrogate model combined with sensitivity analysis to identify and predict U-10Mo microstructure development during thermomechanical processing. Over 200 simulations were collected using physics-based microstructure models covering a wide range of thermomechanical processing routes and initial alloy grain features. Based on the sensitivity analysis, we determined that an increase in rolling reduction percentage at each processing pass has the strongest effect in reducing the grain size. Multi-pass rolling and annealing can significantly improve recrystallization regardless of the reduction percentage. With a volume fraction below 2%, uranium carbide particles were found to have marginal effects on the average grain size and distribution. The proposed stratified stacking ensemble surrogate predicts the U-10Mo grain size with a mean square error four times smaller than a standard single deep neural network. At the same time, with a significant speedup (1000×) compared to the physics-based model, the machine learning surrogate shows good potential for U-10Mo fabrication process optimization.

36 MATERIALS SCIENCE↗

Simultaneous multi-wavelength campaign on PKS 2005-489 in a high state

The high-frequency peaked BL Lac object PKS 2005-489 was the target of amulti-wavelength campaignwith simultaneous observations in the TeV γ-ray (H.E.S.S.), GeV γ-ray (Fermi/LAT), X-ray (RXTE, Swift), UV (Swift) and optical (ATOM, Swift) bands. This campaign was carried out during a high flux state in the synchrotron regime. The flux in the optical and X-ray bands reached the level of the historical maxima. The hard GeV spectrum observed with Fermi/LAT connects well to the very high energy (VHE, E> 100 GeV) spectrum measured with H.E.S.S. with a peak energy between ~ 5 and 500 GeV. Compared to observations with contemporaneous coverage in the VHE and X-ray bands in 2004, the X-ray flux was ~ 50 times higher during the 2009 campaign while the TeV γ-ray flux shows marginal variation over the years. The spectral energy distribution during this multi-wavelength campaign was fit by a one zone synchrotron self-Compton model with a well determined cutoff in X-rays. The parameters of a one zone SSC model are inconsistent with variability time scales. The variability behaviour over years with the large changes in synchrotron emission and small changes in the inverse Compton emission does not warrant an interpretation within a one-zone SSC model despite an apparently satisfying fit to the broadband data in 2009.

79 ASTRONOMY AND ASTROPHYSICS↗

Constraining ΛCDM cosmological parameters with Einstein Telescope mock data

ABSTRACT We investigate the capability of Einstein Telescope to constrain the cosmological parameters of the non-flat ΛCDM cosmological model. Two types of mock data sets are considered depending on whether or not a short gamma-ray burst is detected, and associated with the gravitational wave emitted by binary neutron stars merger, using the THESEUS satellite. Depending on the mock data set, two statistical estimators are applied: one assumes that the redshift is known, while the other marginalizes over it assuming a specific redshift prior distribution. We demonstrate that (i) using mock catalogues collecting gravitational wave signals emitted by binary neutron stars systems to which a short gamma-ray burst has been associated, Einstein Telescope may achieve an accuracy on the cosmological parameters of $\sigma _{H_0}\approx 0.40$ km s−1 Mpc−1, $\sigma _{\Omega _{k,0}}\approx 0.09$, and $\sigma _{\Omega _{\Lambda ,0}}\approx 0.07$; while (ii) using mock catalogues collecting all gravitational wave signals emitted by binary neutron stars systems for which an electromagnetic counterpart has not been detected, Einstein Telescope may achieve an accuracy on the cosmological parameters of $\sigma _{H_0}\approx 0.04$ km s−1 Mpc−1, $\sigma _{\Omega _{k,0}}\approx 0.01$, and $\sigma _{\Omega _{\Lambda ,0}}\approx 0.01$, once the redshift probability distribution of GW events is known from from population synthesis simulations and/or the measure of the tidal deformability parameter. These results show an improvement of a factor 2–75 with respect to earlier results using complementary data sets.

Califano, Matteo↗

Portal dose image prediction for in vivo treatment verification completely based on EPID measurements

A high dosimetric accuracy is required for radiotherapy treatments where IMRT in combination with narrow treatment margins is applied to achieve optimally conformal dose distributions. In order to routinely verify the in vivo fluence delivery (i.e., during the actual patient treatment), our method for predicting portal dose images with a patient in the beam was validated. A unique feature of this method is that it is fully based on calibration measurements with an EPID. The portal dose image (PDI) behind a patient is dependent on the transmission of primary radiation through the patient and a contribution of scattered radiation from the patient. To derive both components, the patient geometry as observed in the planning CT scan is converted into an equivalent homogeneous phantom. A limited set of EPID measurements is required to derive the input parameters of this model. The accuracy of the in vivo PDI prediction was verified using measurements behind phantoms and four prostate cancer patients treated with IMRT. Behind homogeneous slab phantoms, the local differences between measured and predicted PDIs were within 2% inside the field, while behind a lung and a pelvic phantom, the agreement was within 3% or within 3 mm in regions with steep gradients. Outside the fields, the PDIs agreed within 2% of the global dose maximum. Evaluation of the in vivo PDI measurements behind patients showed that, on average, 87% of the pixels inside the field fulfilled the 3% local dose and 3 mm distance-to-agreement criteria. For half of the failing pixels the differences occurred due to changes in patient geometry with respect to the planning CT or due to beam attenuation by the treatment couch that was not accounted for. A fully EPID-based method for predicting portal dose images using the planning CT scan has been implemented and validated for phantoms and clinical patients.

61 RADIATION PROTECTION AND DOSIMETRY↗