Search NASA⌕ Search

SEARCH · Search NASA

Results for “convolution”

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 793 records · Page 44

SU-E-T-02: 90Y Microspheres Dosimetry Calculation with Voxel-S-Value Method: A Simple Use in the Clinic

Purpose: To present a simple and feasible method of voxel-S-value (VSV) dosimetry calculation for daily clinical use in radioembolization (RE) with {sup 90}Y microspheres. Dose distributions are obtained and visualized over CT images. Methods: Spatial dose distributions and dose in liver and tumor are calculated for RE patients treated with Sirtex Medical miscrospheres at our center. Data obtained from the previous simulation of treatment were the basis for calculations: Tc-99m maggregated albumin SPECT-CT study in a gammacamera (Infinia, General Electric Healthcare.). Attenuation correction and ordered-subsets expectation maximization (OSEM) algorithm were applied.For VSV calculations, both SPECT and CT were exported from the gammacamera workstation and registered with the radiotherapy treatment planning system (Eclipse, Varian Medical systems). Convolution of activity matrix and local dose deposition kernel (S values) was implemented with an in-house developed software based on Python code. The kernel was downloaded from www.medphys.it. Final dose distribution was evaluated with the free software Dicompyler. Results: Liver mean dose is consistent with Partition method calculations (accepted as a good standard). Tumor dose has not been evaluated due to the high dependence on its contouring. Small lesion size, hot spots in health tissue and blurred limits can affect a lot the dose distribution in tumors. Extra work includes: export and import of images and other dicom files, create and calculate a dummy plan of external radiotherapy, convolution calculation and evaluation of the dose distribution with dicompyler. Total time spent is less than 2 hours. Conclusion: VSV calculations do not require any extra appointment or any uncomfortable process for patient. The total process is short enough to carry it out the same day of simulation and to contribute to prescription decisions prior to treatment. Three-dimensional dose knowledge provides much more information than other methods of dose calculation usually applied in the clinic.

60 APPLIED LIFE SCIENCES↗

SU-E-T-456: Impact of Dose Calculation Algorithms On Biologically Optimized VMAT Plans for Esophageal Cancer

Purpose: To evaluate the impact of dose calculation algorithm on the dose distribution of biologically optimized Volumatric Modulated Arc Therapy (VMAT) plans for Esophgeal cancer. Methods: Eighteen retrospectively treated patients with carcinoma esophagus were studied. VMAT plans were optimized using biological objectives in Monaco (5.0) TPS for 6MV photon beam (Elekta Infinity). These plans were calculated for final dose using Monte Carlo (MC), Collapsed Cone Convolution (CCC) & Pencil Beam Convolution (PBC) algorithms from Monaco and Oncentra Masterplan TPS. A dose grid of 2mm was used for all algorithms and 1% per plan uncertainty maintained for MC calculation. MC based calculations were considered as the reference for CCC & PBC. Dose volume histogram (DVH) indices (D95, D98, D50 etc) of Target (PTV) and critical structures were compared to study the impact of all three algorithms. Results: Beam models were consistent with measured data. The mean difference observed in reference with MC calculation for D98, D95, D50 & D2 of PTV were 0.37%, −0.21%, 1.51% & 1.18% respectively for CCC and 3.28%, 2.75%, 3.61% & 3.08% for PBC. Heart D25 mean difference was 4.94% & 11.21% for CCC and PBC respectively. Lung Dmean mean difference was 1.5% (CCC) and 4.1% (PBC). Spinal cord D2 mean difference was 2.35% (CCC) and 3.98% (PBC). Similar differences were observed for liver and kidneys. The overall mean difference found for target and critical structures was 0.71±1.52%, 2.71±3.10% for CCC and 3.18±1.55%, 6.61±5.1% for PBC respectively. Conclusion: We observed a significant overestimate of dose distribution by CCC and PBC as compared to MC. The dose prediction of CCC is closer (<3%) to MC than that of PBC. This can be attributed to poor performance of CCC and PBC in inhomogeneous regions around esophagus. CCC can be considered as an alternate in the absence of MC algorithm.

60 APPLIED LIFE SCIENCES↗

SU-F-BRA-15: Physical Aspects and Clinical Applications of Post-Treatment Yttrium-90 PET-Based Dosimetry

Purpose: The accuracy of PET-based post-treatment dosimetry of yttrium-90 microspheres has been improving over the past decade and is now at a stage, permitting volumetric dose-outcome studies. We outline the recent advances and identify the physical limitations to the accuracy of the dose calculations. Methods: Convolution of the measured PET activity density distribution with a pre-calculated voxel-dose-kernel (VDK) is the most widely used method for dose reconstruction. Therefore, accurate knowledge of the beta+ branching ratio as well as the micro-dosimetric characteristics of electron interactions within the microsphere is essential for the computation of the dose kernel. We implement a model of the microspheres used in our clinic to calculate the modified electron energy spectrum at the microsphere’s surface and to determine the impact of self-shielding on the dose reconstruction. The three-dimensional dose distributions obtained for 10 patients treated with radio-embolization with yttrium-90 microspheres are evaluated and various DVH markers are investigated for correlation with outcome. Results: The methods for reducing the overall computation uncertainty are systematically outlined in this presentation. Since the latest experimental data on yttrium-90 beta+ branching ratio has a relative uncertainty of 1.5%, all contributing factors derived from Monte Carlo simulations must be brought to a sub 1% level. The self-shielding within the microspheres is found to be responsible for up to 6% reduction of the reconstructed dose in low-gradient regions and must be taken into account. The contribution of trace amounts of other beta+ emitters introduced during the manufacturing process is also discussed. Conclusion: The accelerating pace of clinical adoption of PET-based post-treatment dosimetry is mainly due to advances in both quantitative PET imaging and physical models of dose deposition. We show that the overall physical dose uncertainty in the convolution step can be further reduced, thus making ongoing multi-institutional dose-outcome studies even more reliable.

60 APPLIED LIFE SCIENCES↗

TU-AB-BRC-06: Dose Calculation in Curved Space

Purpose: Microbeam Radiation Therapy is a preclinical method in radiation oncology that modulates radiation fields on a micrometre scale. Dose calculation is challenging due to arising dose gradients and therapeutically important dose ranges. Monte Carlo (MC) simulations, often used as gold standard, are computationally expensive and hence too slow for the optimisation of treatment parameters in future clinical applications. On the other hand, conventional kernel based dose calculation leads to inaccurate results close to material interfaces. The purpose of this work is to overcome these inaccuracies while keeping computation times low. Methods: A point kernel superposition algorithm is modified to account for tissue inhomogeneities. Instead of conventional ray tracing approaches, methods from differential geometry are applied and the space around the primary photon interaction is locally warped. The performance of this approach is compared to MC simulations and a simple convolution algorithm (CA) for two different phantoms and photon spectra. Results: While peak doses of all dose calculation methods agreed within less than 4% deviations, the proposed approach surpassed a simple convolution algorithm in accuracy by a factor of up to 3 in the scatter dose. In a treatment geometry similar to possible future clinical situations differences between Monte Carlo and the differential geometry algorithm were less than 3%. At the same time the calculation time did not exceed 15 minutes. Conclusion: With the developed method it was possible to improve the dose calculation based on the CA method with respect to accuracy especially at sharp tissue boundaries. While the calculation is more extensive than for the CA method and depends on field size, the typical calculation time for a 20×20 mm{sup 2} field on a 3.4 GHz and 8 GByte RAM processor remained below 15 minutes. Parallelisation and optimisation of the algorithm could lead to further significant calculation time reductions.

60 APPLIED LIFE SCIENCES↗

TH-CD-206-05: Machine-Learning Based Segmentation of Organs at Risks for Head and Neck Radiotherapy Planning

Purpose: Accurate and efficient delineation of tumor target and organs-at-risks is essential for the success of radiotherapy. In reality, despite of decades of intense research efforts, auto-segmentation has not yet become clinical practice. In this study, we present, for the first time, a deep learning-based classification algorithm for autonomous segmentation in head and neck (HaN) treatment planning. Methods: Fifteen HN datasets of CT, MR and PET images with manual annotation of organs-at-risk (OARs) including spinal cord, brainstem, optic nerves, chiasm, eyes, mandible, tongue, parotid glands were collected and saved in a library of plans. We also have ten super-resolution MR images of the tongue area, where the genioglossus and inferior longitudinalis tongue muscles are defined as organs of interest. We applied the concepts of random forest- and deep learning-based object classification for automated image annotation with the aim of using machine learning to facilitate head and neck radiotherapy planning process. In this new paradigm of segmentation, random forests were used for landmark-assisted segmentation of super-resolution MR images. Alternatively to auto-segmentation with random forest-based landmark detection, deep convolutional neural networks were developed for voxel-wise segmentation of OARs in single and multi-modal images. The network consisted of three pairs of convolution and pooing layer, one RuLU layer and a softmax layer. Results: We present a comprehensive study on using machine learning concepts for auto-segmentation of OARs and tongue muscles for the HaN radiotherapy planning. An accuracy of 81.8% in terms of Dice coefficient was achieved for segmentation of genioglossus and inferior longitudinalis tongue muscles. Preliminary results of OARs regimentation also indicate that deep-learning afforded an unprecedented opportunities to improve the accuracy and robustness of radiotherapy planning. Conclusion: A novel machine learning framework has been developed for image annotation and structure segmentation. Our results indicate the great potential of deep learning in radiotherapy treatment planning.

60 APPLIED LIFE SCIENCES↗

Neuromorphic computing for temporal scientific data classification

In this work, we apply a spiking neural network model and an associated memristive neuromorphic implementation to an application in classifying temporal scientific data. We demonstrate that the spiking neural network model achieves comparable results to a previously reported convolutional neural network model, with significantly fewer neurons and synapses required.

Schuman, Catherine↗

A sequential DNN based Baseline Energy Prediction Framework with Long term Error Mitigation

In this paper,we aim to present a novel sequential framework containing two deep network architectures for baseline energy prediction, applied to a simulated commercial building dataset. We have observed that this sequential framework helps to improve long term prediction accuracy, there by mitigating prediction error accumulation over time. The proposed framework utilizes convolution layers to extract features from the input data space without changing spatial relations between variables. These features are memorized by tensor train based gated recurrent units for an accurate long-term prediction. Although, architecture comprising the amalgamation of convolution layers and memory cell have shown promising results in domains such as social media analysis, language modeling, video frame prediction and image recognition, this paper extends its scope to the context of energy applications. Furthermore, the addition of tensor train based gated recurrent units is motivated by the necessity of computational time reduction during the training process, there by making this framework more suitable for future in-field deployment. Lastly, results show the current framework outperforms other existing machine learning based methods, in both short-term and long-term prediction categories.

Chakraborty, Indrasis↗

Effects of Radiation-Induced Defects on Corrosion

The next generation of nuclear reactors will expose materials to conditions that, in some cases, are even more extreme than those in current fission reactors, inevitably leading to new materials science challenges. Radiation-induced damage and corrosion are two key phenomena that must be understood both independently and synergistically, but their interactions are often convoluted. In the light water reactor community, a tremendous amount of work has been done to illuminate irradiation-corrosion effects, and similar efforts are under way for heavy liquid metal and molten salt environments. While certain effects, such as radiolysis and irradiation-assisted stress corrosion cracking, are reasonably well established, the basic science of how irradiation-induced defects in the base material and the corrosion layer influence the corrosion process still presents many unanswered questions. In this review, we summarize the work that has been done to understand these coupled extremes, highlight the complex nature of this problem, and identify key knowledge gaps.

36 MATERIALS SCIENCE↗

pnnl/grid_prediction

Two datadriven predictive approaches, namely, {\em Koopman Operator Theoretic (KOT)-based} model and {\em Graph Neural Network (GNN)-based model}, to enable effective power system state predictions. The KOT-based approaches (Robust DMD, deepDMD) capture the power system evolution as a linear dynamical system on an abstract space. The GNNs model the spatio-temporal correlations using graph convolutional network and are called Spatio-Temporal Graph Convolutional Network (STGCN). These predictive models are trained, tested and compared rigorously based on their predictions of frequencies in the IEEE 68 bus system when subjected to a disturbance. GridSTAGE framework developed at Pacific Northwest National Laboratory is leveraged to generate multiple datasets (in the form of PMU measurements) for training and testing by strategically creating load changes across the spatial locations of the network

Nandanoor, Sai Pushpak↗

Green's function fast multipole method for continuum mechanics (SM-FMM)

Solid Mechanics Fast Multipole Method based on elastic Green's function and accelerated using FFTs The code calculates the mechanical fields (stress and strain) for a heterogeneous elasto-plastic material unit cell under quasi-static conditions (no dynamic effects). The fields are calculated using a Green's function method, where the strain field is given by a discrete convolution of Green's operator with an auxiliary stress field. The convolution is calculated using the fast multipole method.

Zecevic, Miroslav↗

torch-einshard v1.0

torch-einshard is a Python library for describing local and distributed PyTorch tensor computations with compact, einsum-like notation. Its expressions name logical axes, specify how they are sharded across a PyTorch DeviceMesh, and represent partial reductions. The library automatically performs contractions, permutations, reshaping, splitting, gathering, reduction, reduce-scatter, and repartitioning while preserving autograd. Additional features include sharding-aware FFTs, tensor rolls, halo exchange, sliding windows, 1D–3D convolutions, uneven-shard handling, parameter initialization and gradient management, and cost-based execution planning. It is designed for scientific machine learning and large-model workloads, including tensor-, sequence-, and spatial-parallel MLPs, attention, convolutions, and spectral operations. Compared with manually combining torch.einsum and distributed collectives, torch-einshard expresses both the mathematical operation and data placement in one readable formula. This reduces boilerplate and synchronization errors, keeps forward and backward communication consistent, and allows the library to select optimized collective strategies without changing model code.

Morozov, Dmitriy [Lawrence Berkeley National Labor↗

Exploring Randomly Wired Neural Networks for Climate Model Emulation

Exploring the climate impacts of various anthropogenic emissions scenarios is key to making informed decisions for climate change mitigation and adaptation. State-of-the-art Earth system models can provide detailed insight into these impacts but have a large associated computational cost on a per-scenario basis. This large computational burden has driven recent interest in developing cheap machine learning models for the task of climate model emulation. In this paper, we explore the efficacy of randomly wired neural networks for this task. We describe how they can be constructed and compare them with their standard feedforward counterparts using the ClimateBench dataset. Specifically, we replace the serially connected dense layers in multilayer perceptrons, convolutional neural networks, and convolutional long short-term memory networks with randomly wired dense layers and assess the impact on model performance for models with 1 million and 10 million parameters. We find that models with less-complex architectures see the greatest performance improvement with the addition of random wiring (up to 30.4% for multilayer perceptrons). Furthermore, of 24 different model architecture, parameter count, and prediction task combinations, only one had a statistically significant performance deficit in randomly wired networks relative to their standard counterparts, with 14 cases showing statistically significant improvement. We also find no significant difference in prediction speed between networks with standard feedforward dense layers and those with randomly wired layers. These findings indicate that randomly wired neural networks may be suitable direct replacements for traditional dense layers in many standard models.

54 ENVIRONMENTAL SCIENCES↗

Neural Networks to Find the Optimal Forcing for Offsetting the Anthropogenic Climate Change Effects

Abstract Of great relevance to climate engineering is the systematic relationship between the radiative forcing to the climate system and the response of the system, a relationship often represented by the linear response function (LRF) of the system. However, estimating the LRF often becomes an ill-posed inverse problem due to high-dimensionality and nonunique relationships between the forcing and response. Recent advances in machine learning make it possible to address the ill-posed inverse problem through regularization and sparse system fitting. Here, we develop a convolutional neural network (CNN) for regularized inversion. The CNN is trained using the surface temperature responses from a set of Green’s function perturbation experiments as imagery input data together with data sample densification. The resulting CNN model can infer the forcing pattern responsible for the temperature response from out-of-sample forcing scenarios. This promising proof of concept suggests a possible strategy for estimating the optimal forcing to negate certain undesirable effects of climate change. The limited success of this effort underscores the challenges of solving an inverse problem for a climate system with inherent nonlinearity. Significance Statement Predicting the climate response for a given climate forcing is a direct problem, while inferring the forcing for a given desired climate response is often an inverse, ill-posed, problem, posing a new challenge to the climate community. This study makes the first attempt to infer the radiative forcing for a given target pattern of global surface temperature response using a deep learning approach. The resulting deeply trained convolutional neural network inversion model shows promise in capturing the forcing pattern corresponding to a given surface temperature response, with a significant implication on the design of an optimal solar radiation management strategy for curbing global warming. This study also highlights the technical challenges that future research should prioritize in seeking feasible solutions to the inverse climate problem.

Ren, Huiying↗

Evaluating the Trustworthiness of Explainable Artificial Intelligence (XAI) Methods Applied to Regression Predictions of Arctic Sea Ice Motion

Abstract Recent advances in explainable artificial intelligence (XAI) methods show promise for understanding predictions made by machine learning (ML) models. XAI explains how the input features are relevant or important for the model predictions. We train linear regression (LR) and convolutional neural network (CNN) models to make 1-day predictions of sea ice velocity in the Arctic from inputs of present-day wind velocity and previous-day ice velocity and concentration. We apply XAI methods to the CNN and compare explanations to variance explained by LR. We confirm the feasibility of using a novel XAI method [i.e., global layerwise relevance propagation (LRP)] to understand ML model predictions of sea ice motion by comparing it to established techniques. We investigate a suite of linear, perturbation-based, and propagation-based XAI methods in both local and global forms. Outputs from different explainability methods are generally consistent in showing that wind speed is the input feature with the highest contribution to ML predictions of ice motion, and we discuss inconsistencies in the spatial variability of the explanations. Additionally, we show that the CNN relies on both linear and nonlinear relationships between the inputs and uses nonlocal information to make predictions. LRP shows that wind speed over land is highly relevant for predicting ice motion offshore. This provides a framework to show how knowledge of environmental variables (i.e., wind) on land could be useful for predicting other properties (i.e., sea ice velocity) elsewhere. Significance Statement Explainable artificial intelligence (XAI) is useful for understanding predictions made by machine learning models. Our research establishes trustability in a novel implementation of an explainable AI method known as layerwise relevance propagation for Earth science applications. To do this, we provide a comparative evaluation of a suite of explainable AI methods applied to machine learning models that make 1-day predictions of Arctic sea ice velocity. We use explainable AI outputs to understand how the input features are used by the machine learning to predict ice motion. Additionally, we show that a convolutional neural network uses nonlinear and nonlocal information in making its predictions. We take advantage of the nonlocality to investigate the extent to which knowledge of wind on land is useful for predicting sea ice velocity elsewhere.

Hoffman, Lauren [Scripps Institution of Oceanograp↗

On the three-dimensional spatial correlations of curved dislocation systems

Abstract Coarse-grained descriptions of dislocation motion in crystalline metals inherently represent a loss of information regarding dislocation-dislocation interactions. In the present work, we consider a coarse-graining framework capable of re-capturing these interactions by means of the dislocation-dislocation correlation functions. The framework depends on a convolution length to define slip-system-specific dislocation densities. Following a statistical definition of this coarse-graining process, we define a spatial correlation function which will allow the arrangement of the discrete line system at two points—and thus the strength of their interactions at short range—to be recaptured into a mean field description of dislocation dynamics. Through a statistical homogeneity argument, we present a method of evaluating this correlation function from discrete dislocation dynamics simulations. Finally, results of this evaluation are shown in the form of the correlation of dislocation densities on the same slip-system. These correlation functions are seen to depend weakly on plastic strain, and in turn, the dislocation density, but are seen to depend strongly on the convolution length. Implications of these correlation functions in regard to continuum dislocation dynamics as well as future directions of investigation are also discussed.

Anderson, Joseph Pierre (ORCID:0000000329760903)↗

Deep Neural Network Algorithm for CMC Microstructure Characterization and Variability Quantification

Microstructure characterization and variability quantification are crucial for understanding ceramic matrix composites (CMCs) mechanical behavior and deformation mechanisms across length scales. Traditionally, analyses of the micrographs obtained from microscopy are labor-intensive. However, with the vast improvement in computer vision (CV) and deep learning (DL), an automated algorithm can be designed to extract essential microstructure variability from micrographs which can then be used to construct a statistically representative volume element (SRVE). The DL-based algorithm spans the taxonomy of microstructure analyses, including semantic segmentation of microstructure constituents, secondary phases, matrix/fiber interface, and defects, and quantifying the microstructure variability in terms of probability distributions. In this work, C/SiNC and SiC/SiNC CMCs microstructures are semantically segmented through a deep convolutional neural network, followed by variability quantification through the implementation of a fully connected regression layer, hence forming a deep regression network. The deep regression network operates in a feedforward regime, in which the neuron output signal traverses through the network in a unidirectional manner. The weight tensor associated with each layer is updated through a backpropagation stochastic gradient descent approach. The input gray-scale image obtained through in-house scanning electron microscope and confocal microscope micrographs is augmented through affine transformations to increase the training set size, which is then processed through four strided convolutional layers. This compresses the image resolution by half at each layer while increasing the image depth by applying different filters (image encoding). The class activation maps (CAMs) corresponding to the applied filters highlight the key architectural features and assist with the semantic segmentation of the microstructure.

Hamza, Mohamed H.↗

Inverse Analysis with Variational Autoencoders: A Comparison of Shallow and Deep Networks

Inverse problems are applied to determine unknown properties by matching observational data with a physical model that often takes many parameters as input. To overcome the underconstrained nature of inverse problems and achieve good performance, an approach is presented involving regularization with a technique known as a variational autoencoder (VAE), which is trained to map a high-dimensional parameter space with a complex structure to a low-dimensional latent space with a simple structure. We apply this approach to unconditioned realizations of the parameters (heterogeneous hydraulic fields) for a hydrogeological inverse problem. Two types of hydraulic conductivity fields are used to evaluate the characterization for the different levels of heterogeneity complexity of the physical inputs. This approach keeps the computational cost of generating the training data low. The reason is unconditioned realizations neither rely on the observational data used to perform the inverse analysis nor require any groundwater flow model forward runs. In addition, this approach applies regularization on a low-dimensional latent space from the VAE and increases optimization efficiency through automatic differentiation. Furthermore, two different neural network (NN) structures are tested for their utility in using the VAE for inverse analysis. The performance of a deep, convolutional neural network strongly depends on the dimensionality of the latent space, which requires tuning. In contrast, a shallow, dense neural network provides consistently accurate characterization without tuning. Furthermore, our approach evaluates the advantages of the shallow, dense neural network over the deep, convolutional one and enables future application to a wide range of inverse problems.

97 MATHEMATICS AND COMPUTING↗

Applying Machine Learning and Bayesian Inference to Identify and Locate Moving Anthropogenic Sources Using Distributed Acoustic Sensing Data

Distributed acoustic sensing (DAS) systems, which use existing telecommunication fibers, offer high‐resolution capabilities ideal for recording anthropogenic sources. However, the complexity of urban environments and the large amount of data recorded by DAS require automated methods to efficiently detect and categorize anthropogenic sources. Here, we evaluate how well three machine learning models (k‐nearest neighbor [k‐NN], convolutional neural networks, and recurrent‐convolutional neural networks) can identify various anthropogenic sources recorded by DAS. Our findings reveal that both k‐NN and neural network methods perform well in high signal‐to‐noise ratio (SNR) settings. However, their accuracy decreases at SNRs <4. We also use Kalman filtering, a form of Bayesian inference, on backprojected locations of these sources to recover locations that generally fall within standard smartphone Global Positioning System errors. By combining machine learning and Kalman filter results, we calculate a multidimensional model of moving anthropogenic sources. These results demonstrate the potential of DAS data in urban seismology for accurately identifying and locating such sources. Depending on the research objectives, these sources can be further studied or filtered out to improve the quality of seismic data for earthquake studies. Such methods provide a valuable tool for urban seismology and seismic hazard analysis.

Luckie, Thomas William [Sandia National Laboratori↗