Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Task Performance and Analysis”

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 271 records · Page 15
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

Model Assumptions and Data Characteristics: Impacts on Domain Adaptation in Building Segmentation

Studies on domain adaptation (DA) for remote sensing (RS) imagery analysis lack consistency in selection and description of evaluation scenarios. Without properly characterizing datasets, model assumptions, and evaluation scenarios, it is difficult to objectively compare DA methods and reach conclusions about their suitability across different applications. With this motivation, this work seeks to empirically assess to which extent the interaction between data characteristics and model assumptions influences the effectiveness of DA methods. Using the widely explored task of building footprint segmentation as a case study, we perform a large-scale study across over 200 DA scenarios that include variations across view angles, areas observed, and sensors used for data acquisition. Rather than adopting different model architectures or optimization criteria, we contrast the performances of two DA methods based on adversarial learning that differ only in their assumptions about source and target domains. Informed by metadata and data characteristics unveiled using traditional computer vision (CV) techniques as well as pretrained deep models, we provide a detailed meta-analysis of experiments highlighting the importance of accurately considering data assumptions for DA in RS segmentation tasks. As demonstrated by a “cherry-picking” exercise, different claims regarding which model is best could be made by selecting different subsets of evaluation scenarios. While well-calibrated assumptions can be beneficial, mismatching assumptions can lead to negative biases in DA applications. Furthermore, this study intends to motivate the community toward more consistent evaluation protocols while providing recommendations and insights toward creating novel benchmark datasets, documenting data characteristics, application-specific knowledge, and model assumptions.

42 ENGINEERING↗
DOE OSTI2016

MO-DE-207A-01: Impact of Statistical Weights On Detection of Low-Contrast Details in Model-Based Iterative CT Reconstruction

Purpose: Penalized-weighted least-square reconstruction has become an important research topic in CT, to reduce dose without affecting image quality. Two components impact image quality in this reconstruction: the statistical weights and the use of an edge-preserving penalty term. We are interested in assessing the influence of statistical weights on their own, without the edge-preserving feature. Methods: The influence of statistical weights on image quality was assessed in terms of low-contrast detail detection using LROC analysis. The task amounted to detect and localize a 6-mm lesion with random contrast inside the FORBILD head phantom. A two-alternative forced-choice experiment was used with two human observers performing the task. Reconstructions without and with statistical weights were compared, both using the same quadratic penalty term. The beam energy was set to 30keV to amplify spatial differences in attenuation and thereby the role of statistical weights. A fan-beam data acquisition geometry was used. Results: Visual inspection of images clearly showed a difference in noise between the two reconstructions methods. As expected, the reconstruction without statistical weights exhibited noise streaks. The other reconstruction appeared better in this aspect, but presented other disturbing noise patterns and artifacts induced by the weights. The LROC analysis yield the following 95-percent confidence interval for the difference in reader-averaged AUC (reconstruction without weights minus reconstruction with weights): [0.0026,0.0599]. The mean AUC value was 0.9094. Conclusion: We have investigated the impact of statistical weights without the use of edge-preserving penalty in penalized weighted least-square reconstruction. A decrease rather than increase in image quality was observed when using statistical weights. Thus, the observers were better able to cope with the noise streaks than the noise patterns and artifacts induced by the statistical weights. It may be that different results would be obtained if the penalty term was used with a pixel-dependent weight. F Noo receives research support from Siemens Healthcare GmbH.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2026

Transfer learning for analysis of collective and non-collective Thomson scattering spectra

Thomson scattering (TS) diagnostics provide reliable, minimally perturbative measurements of fundamental plasma parameters, such as electron density (⁠n e ) and electron temperature (⁠T e ⁠). Deep neural networks can provide accurate estimates of ⁠n e and T e when conventional fitting algorithms may fail, such as when TS spectra are dominated by noise, or when fast analysis is required for real-time operation. Although deep neural networks typically require large training sets, transfer learning can improve model performance on a target task with limited data by leveraging pre-trained models from related source tasks, where select hidden layers are further trained using target data. We present five architecturally diverse deep neural networks, pre-trained on synthetic TS data and adapted for experimentally measured TS data, to evaluate the efficacy of transfer learning in estimating n e and T e in both the collective and non-collective scattering regimes. We evaluate errors in n e and T e estimates as a function of training set size for models trained with and without transfer learning, and we observe decreases in model error from transfer learning when the training set contains ≲ 200 experimentally measured spectra.

Artificial neural networks↗
DOE OSTI2018

Efficient, Parallel At-scale Correlation Analysis for Atom Probe Tomography on Hybrid Architectures

Atom probe tomography (APT) is a material probing technique that has undergone dramatic improvements in its capability to map individual atoms within a material sample resulting in data files with hundreds of millions of atoms. Understanding the nano-structural features hidden in these massive amounts of atomic data is a crucial analysis task for materials scientists. However, fast analysis capabilities for large APT workloads remains a critical bottleneck. In this paper, we present the design, implementation and detailed performance evaluations of a parallel software capable of efficiently performing extremely time-consuming correlation analyses of massive high density APT data. Starting with shared memory implementations to motivate our design choices, we extend the implementation to hybrid architectures keeping realistic APT workloads in mind. Detailed performance analyses of three different parallel implementations of the software are supported by empirical results on a Cray XC30 and a Cray XC40 architecture. Its usefulness is demonstrated by reducing the turnaround time of an end-to-end APT correlation analysis on 100 millions atoms by three orders of magnitude using 2048 MPI ranks on 1024 nodes (24 cores per node) of a Cray XC30. The software reported here equips material scientists for the first time with a high-speed scalable capability for efficient and timely analyses of massive APT data.

Lu, Hao↗
DOE OSTI2025

Accurate and Fast Anomaly Detection in Additive Composite-Based Manufacturing using Thermal Cameras

Today, large-scale additive manufacturing with plastics and composite materials requires continuous monitoring by experienced staff to prevent, detect and correct anomalous events affecting the performance of the printed part. We address the complexity of this demanding task by designing a camera-based anomaly detection system utilizing probabilistic principal component analysis (PPCA). This is a machine learning technique is trained with thermal images collected during normal operation of the large-scale printer (Cincinnati BAAM). This technique is advantageous for practical applications as there is no need to artificially introduce anomalous conditions into model training. During deployment, we challenge this model by introducing deliberate variations of the extruder speed. We reduce extrusion speed to a lower level, between 70 and 95% of the nominal value to collected test images. Our results show that images are easily identified as anomalous for extruder speeds at or below 85% of the nominal speed, meaning that an anomalous reduction of the material deposition rate can be detected within seconds of its onset. We show that our results are robust to (a) camera-to-camera variability and (b) print-to-print variability.

Pike, John [ORNL]↗
DOE OSTI2015

Transition of Iodine Analysis to Accelerator Mass Spectrometry

The goal of the project, New Paradigms for Isotope Ratio Mass Spectrometry: Raising the Scientific Profile and Improved Performance for Accelerator Mass Spectrometry (AMS) and Thermal Ionization Mass Spectrometry (TIMS), is to ensure that the ongoing isotope ratio determination capability within the U.S. Department of Energy complex is the world’s best for application to nonproliferation. This report spells out the progress of Task 4, Transition of TIMS to AMS for Iodine Analysis, of the larger project. The subtasks under Task 4 and the accomplishments throughout the three year project life cycle are presented in this report. Progress was made in optimization of chemical extraction, determination of a detection limit for 127Iodine, production of standard materials for AMS analysis quality assurance, facilitation of knowledge exchange with respect to analyzing iodine on an AMS, cross comparison with a world-leading AMS laboratory, supercritical fluid extraction of iodine for AMS analysis and electrodeposition of seawater as a direct method of preparation for iodine analysis by AMS--all with the goal of minimizing the time required to stand up an AMS capability for iodine analysis of exposed air filters at INL. An effective extraction method has been developed and demonstrated for iodine analysis of exposed air filters. Innovative techniques to accomplish the cathode preparation for AMS analysis were developed and demonstrated and published. The known gap of a lack of available materials for reference standards in the analysis of iodine by AMS was filled by the preparation of homogenous materials that were calibrated against NIST materials. A minimum limit on the amount of abundant isotope in a sample was determined for AMS analysis. The knowledge exchange occurred with fantastic success. Scientists engaged the international AMS community at conferences, as well as in their laboratories for collaborative work. The supercritical fluid extraction work has positive data, but is not a replacement for leaching yet. The added seawater work has led to a good method and a possible publication. The focus of this project was to minimize the time to stand up the AMS capability, by having all the preparation and supporting functions worked out ahead of the instrument arrival. Due to all the preparatory work and its success, the instrument was delivered and turned over to the INL in February 2015. Since then, INL scientists have been successfully vetting the capabilities and accomplishing their own measurements that agree well with the leading laboratories of the world for iodine analysis by AMS. Initially, all AMS data was gathered on other laboratories instruments, but during the last six months data for this project has come from INL’s AMS.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2016

Integral Reactor Containment Condensation Model and Experimental Validation

This NEUP funded project, NEUP 12-3630, is for experimental, numerical and analytical studies on high-pressure steam condensation phenomena in a steel containment vessel connected to a water cooling tank, carried out at Oregon State University (OrSU) and the University of Wisconsin at Madison (UW-Madison). In the three years of investigation duration, following the original proposal, the planned tasks have been completed: (1) Performed a scaling study for the full pressure test facility applicable to the reference design for the condensation heat transfer process during design basis accidents (DBAs), modified the existing test facility to route the steady-state secondary steam flow into the high pressure containment for controllable condensation tests, and extended the operations at negative gage pressure conditions (OrSU). (2) Conducted a series of DBA and quasi-steady experiments using the full pressure test facility to provide a reliable high pressure condensation database (OrSU). (3) Analyzed experimental data and evaluated condensation model for the experimental conditions, and predicted the prototypic containment performance under accidental conditions (UW-Madison). A film flow model was developed for the scaling analysis, and the results suggest that the 1/3 scaled test facility covers large portion of laminar film flow, leading to a lower average heat transfer coefficient comparing to the prototypic value. Although it is conservative in reactor safety analysis, the significant reduction of heat transfer coefficient (50%) could under estimate the prototypic condensation heat transfer rate, resulting in inaccurate prediction of the decay heat removal capability. Further investigation is thus needed to quantify the scaling distortion for safety analysis code validation. Experimental investigations were performed in the existing MASLWR test facility at OrST with minor modifications. A total of 13 containment condensation tests were conducted for pressure ranging from 4 to 21 bar with three different static inventories of non-condensable gas. Condensation and heat transfer rates were evaluated employing several methods, notably from measured temperature gradients in the HTP as well as measured condensate formation rates. A detailed mass and energy accounting was used to assess the various measurement methods and to support simplifying assumptions required for the analysis. Condensation heat fluxes and heat transfer coefficients are calculated and presented as a function of pressure to satisfy the objectives of this investigation. The major conclusions for those tests are summarized below: (1) In the steam blow-down tests, the initial condensation heat transfer process involves the heating-up of the containment heat transfer plate. An inverse heat conduction model was developed to capture the rapid transient transfer characteristics, and the analysis method is applicable to SMR safety analysis. (2) The average condensation heat transfer coefficients for different pressure conditions and non-condensable gas mass fractions were obtained from the integral test facility, through the measurements of the heat conduction rate across the containment heat transfer plate, and from the water condensation rates measurement based on the total energy balance equation. 15 (3) The test results using the measured HTP wall temperatures are considerably lower than popular condensation models would predict mainly due to the side wall conduction effects in the existing MASLWR integral test facility. The data revealed the detailed heat transfer characteristics of the model containment, important to the SMR safety analysis and the validation of associated evaluation model. However this approach, unlike separate effect tests, cannot isolate the condensation heat transfer coefficient over the containment wall, and therefore is not suitable for the assessment of the condensation heat transfer coefficient against system pressure and noncondensable gas mass fraction. (4) The average condensation heat transfer coefficients measured from the water condensation rates through energy balance analysis are appropriate, however, with considerable uncertainties due to the heat loss and temperature distribution on the containment wall. With the consideration of the side wall conduction effects, the results indicate that the measured heat transfer coefficients in the tests is about 20% lower than the prediction of Dehbi’s correlation, mainly due to the side wall conduction effects. The investigation also indicates an increase in the condensation heat transfer coefficient at high containment pressure conditions, but the uncertainties invoked with this method appear to be substantial. (5) Non-condensable gas in the tests has little effects on the condensation heat transfer at high elevation measurement ports. It does affect the bottom measurements near the water level position. The results suggest that the heavier non-condensable gas is accumulated in the lower portion of the containment due to stratification in the narrow containment space. The overall effects of the non-condensable gas on the heat transfer process should thus be negligible for tall containments of narrow condensation spaces in most SMR designs. Therefore, the previous correlations with noncondensable gas effects are not appropriate to those small SMR containments due to the very poor mixing of steam and non-condensable gas. The MELCOR simulation results agree with the experimental data reasonably well. However, it is observed that the MELCOR overpredicts the heat flux for all analyzed tests. The MELCOR predicts that the heat fluxes for CCT’s approximately range from 30 to 45 kW/m2 whereas the experimental data (averaged) ranges from about 25 to 40 kW/m2. This may be due to the limited availability of liquid film models included in MELCOR. Also, it is believed that due to complex test geometry, measured temperature gradients across the heat transfer plate may have been underestimated and thus the heat flux had been underestimated. The MELCOR model predicts a film thickness on the order of 100 microns, which agrees very well with film flow model developed in this study for scaling analysis. However, the expected differences in film thicknesses for near vacuum and near atmospheric test conditions are not significant. Further study on the behavior of condensate film is expected to refine the simulation results. Possible refinements include but are not limited to, the followings: CFD simulation focusing on the liquid film behavior and benchmarking with experimental analyses for simpler geometries. 16 1 INTRODUCTION This NEUP funded project, NEUP 12-3630, is for experimental, numerical and analytical studies on high-pressure steam condensation phenomena in a steel containment vessel connected to a water cooling tank, carried out at Oregon State University (OrSU) and the University of Wisconsin at Madison (UW-Madison). The experimental results are employed to validate the containment condensation model in reactor containment system safety analysis code for integral SMRs. Such a containment condensation model is important to demonstrate the adequate cooling. In the three years of investigation, following the original proposal, the following planned tasks have been completed: (1) Performed a scaling study for the full pressure test facility applicable to the reference design for the condensation heat transfer process during design basis accidents (DBAs), modified the existing test facility to route the steady-state secondary steam flow into the high pressure containment for controllable condensation tests, and extended the operations at negative gage pressure conditions (OrSU). (2) Conducted a series of DBA and quasi-steady experiments using the full pressure test facility to provide a reliable high pressure condensation database (OrSU). (3) Analyzed experimental data and evaluated condensation model for the experimental conditions, and predicted the prototypic containment performance under accidental conditions (UW-Madison). The results are applicable to integral Small Modular Reactor (SMR) designs, including NuScale, mPower, Westinghouse SMR, Holtec-160 and other integral reactors with small containments of relatively high pressures under accidental conditions. Testing has been conducted at the OrSU laboratory in the existing MASLWR (Multi-Application Small Light Water Reactor) integral test facility sponsored by the US Department of Energy. Its highpressure stainless steel containment model (~2 MPa) is scaled to the NuScale SMR currently under development at NuScale Power, Inc.. Minor modifications to the model containment have been made to control the non-condensable gas fraction and to utilize the secondary loop stable steam flow for condensation testing. UW-Madison has developed a containment condensation model, which leveraged previous validated containment heat transfer work carried out at UW-Madison, and extended the range of applicability of the model to integral SMR designs that utilize containment vessels of high heat transfer efficiencies. In this final report, the research background and literature survey are presented in Chapter 2 and 3, respectively. The test facility description and modifications are summarized in Chapter 4, and the scaling analysis is introduced in Chapter 5. The tests description, procedures, and data analysis are presented in Chapter 6, while the numerical modeling is presented in Chapter 7, followed by a conclusion section in Chapter 8.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2024

Rancor-HUNTER: A Virtual Plant and Operator Environment for Predicting Human Performance

Advances in simulation capabilities to model physical systems have outpaced the development of simulations for humans using those physical systems. There is an argument that the infinite span of potential human behaviors inherently render human modeling more challenging than physical systems. Despite this challenge, the need for modeling humans interacting with these complex systems is paramount. As technologies have improved, many of the failure modes originating from the physical systems have been solved. This means the overall proportion of human errors has increased, such that it is not uncommon to be the primary driver of system failure in modern complex systems. Moreover, technologies such as automated systems may introduce emerging contexts that can cause new, unanticipated modes of human error. Therefore, it is now more important than ever to develop models of human behavior to realize overall system error reductions and achieve established safety margins. To support new and novel concepts of operations for the anticipated wave of advanced nuclear reactor deployments, human factors and human reliability analysis researchers need to develop advanced simulation-based approaches. This talk presents a simulation environment suitable to both collect data and then perform Monte Carlo simulations to evaluate human performance and develop better models of human behavior. Specifically, the Rancor Microworld Simulator models a complex energy production system in a simplified manner. Rancor includes computer-based procedures, which serve as a framework to automatically classify human behaviors without manual, subjective experimenter coding during scenarios. This method supports a detailed level of analysis at the task level. It is feasible for collecting large sample sizes required to develop quantitative modelling elements that have historically challenged traditional full-scope simulator study approaches. Additionally, the other portion of this experimental platform, the Human Unimodel for Nuclear Technology to Enhance Reliability (HUNTER), is presented to show how the collected data can be used to evaluate novel scenarios based on the contextual factors, or performance shaping factors, derived from Rancor simulations. Rancor-HUNTER is being used to predict operator performance with new procedures, such as results from control room modernization or new-build situations. Rancor-HUNTER is also proving a useful surrogate platform to model human performance for other complex systems.

99 GENERAL AND MISCELLANEOUS↗
DOE OSTI2024

Towards Autonomous Experiments by Connecting High Performance Microscopy with High Performance Computing

The digitization of controls, data, and analysis in microscopy is bringing the idea of autonomous microscopes closer to reality than ever before. Automated transmission electron microscopy (TEM) is already fairly routine for some experiments the only require simple repetitive tasks such as imaging biological macromolecules for single particle cryoEM [1], tilt series for electron tomography [2], and movies for crystallography [3]. The vast majority of TEM experiments are conducted completely by human operators who choose the regions of interest, optimize experimental parameters, and make decisions about data quality visually during an experiment. The field is still a long way from having completely autonomous TEMs that can adapt to sample difficulties and tune experimental parameters based on data quality and desired experimental outcomes. Part of the issue is the lack of capability for feeding information learned from on-line, live data analysis back into the on-going experiment [4]. Furthermore, this presentation will discuss current capabilities for large scale data reduction and analysis using high performance computing (i.e. supercomputing) and progress towards developing a true feed-back loop that places data analysis and theory in the experimental loop.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2019

NEUP Final Report: Analysis of Thermal Transients for a Sodium-CO 2 Heat Exchanger

The design of heat exchangers for use in a supercritical carbon dioxide (sCO 2 ) Brayton cycle power converter must provide for acceptable performance for duty cycle events encompassing anticipated transients and postulated accidents. Argonne National Laboratory participated in the NEUP project, “Thermal Hydraulic and Structural Testing and Modeling of Compact Diffusion-Bonded Heat Exchangers for Supercritical CO 2 Brayton Cycle” with the task of conducting a comprehensive analysis of thermal transients for sCO 2 cycle heat exchangers, with emphasis on the sodium-to-CO 2 heat addition heat exchanger. A range of transients, from normal operation to severe accidents, were simulated with the coupled PDC and SAS4A/SASSYS-1 system level dynamic analysis computer codes. For each transient, the calculated change in the heat exchanger wall temperature is determined as a measure of the thermal loading. In addition, an analysis to find an equivalent helium-CO 2 heat exchanger was carried out to facilitate testing with helium rather than sodium on the hot side.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2021

Development of phenomena identification and ranking table for Westinghouse lead fast reactor’s safety

The Westinghouse Lead-cooled Fast Reactor (LFR) is a medium-size, passively safe, economic, Gen-IV nuclear reactor. An important effort within the Westinghouse LFR program is the development of the safety analysis methodology, which comprises computer code development, model development, and experimental testing. A key initial task associated with the development of the safety analysis methodology is the identification of processes and phenomena that affect the plant's capability to meet selected safety performance indicators. This is accomplished through the development of a Phenomena Identification and Ranking Table (PIRT) for selected accident scenarios which, for this specific PIRT effort, included selected postulated design basis accidents and hypothetical beyond design basis accidents in LFRs. This paper describes the role of PIRT in the development of the Westinghouse LFR safety analysis methodology and the process used in the PIRT development. Specifically, the Westinghouse LFR PIRT assessed importance of pertinent phenomena and identified gaps in their knowledge-base by evaluating current modeling capabilities and data available for validation. The ultimate goal was to provide guidance on computer code development and validation efforts and to prioritize testing to support LFR design and licensing. The key phenomena and processes that are deemed highly important for the safety performance indicators, but for which the state of knowledge is low, are presented. The testing program and analyses development are planned to address significant phenomena in the PIRT.

Lead fast reactor↗
DOE OSTI2021

A Computed Tomography Automation Architecture Developed from a Comprehensive Literature Review

This paper presents a literature review on techniques related to the computed tomography procedure that incorporate automation elements in their research investigations or industrial applications. Computed tomography (CT) is a non-destructive testing (NDT) technique in that the imaging and inspection are performed without damaging the sample, allowing for additional or repeated analysis if necessary. The reviewed literature is organized based on the steps associated with a general NDT task in order to define an end-to-end computed tomography automation architecture. The process steps include activities prior to image collection, during the scan, and after the data are collected. It further reviews efforts related to repeating this process based on a previous scan result. By analyzing the multiple existing but disparate efforts found in the literature, we present a framework for fully automating NDT procedures and discuss the remaining technical gaps in the developed framework.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2014

Advanced probabilistic risk analysis using RAVEN and RELAP-7

RAVEN, under the support of the Nuclear Energy Advanced Modeling and Simulation (NEAMS) program [1], is advancing its capability to perform statistical analyses of stochastic dynamic systems. This is aligned with its mission to provide the tools needed by the Risk Informed Safety Margin Characterization (RISMC) path-lead [2] under the Department Of Energy (DOE) Light Water Reactor Sustainability program [3]. In particular this task is focused on the synergetic development with the RELAP-7 [4] code to advance the state of the art on the safety analysis of nuclear power plants (NPP). The investigation of the probabilistic evolution of accident scenarios for a complex system such as a nuclear power plant is not a trivial challenge. The complexity of the system to be modeled leads to demanding computational requirements even to simulate one of the many possible evolutions of an accident scenario (tens of CPU/hour). At the same time, the probabilistic analysis requires thousands of runs to investigate outcomes characterized by low probability and severe consequence (tail problem). The milestone reported in June of 2013 [5] described the capability of RAVEN to implement complex control logic and provide an adequate support for the exploration of the probabilistic space using a Monte Carlo sampling strategy. Unfortunately the Monte Carlo approach is ineffective with a problem of this complexity. In the following year of development, the RAVEN code has been extended with more sophisticated sampling strategies (grids, Latin Hypercube, and adaptive sampling). This milestone report illustrates the effectiveness of those methodologies in performing the assessment of the probability of core damage following the onset of a Station Black Out (SBO) situation in a boiling water reactor (BWR). The first part of the report provides an overview of the available probabilistic analysis capabilities, ranging from the different types of distributions available, possible sampling strategies, and post processing analysis capabilities. The first part of the report provides an extensive description of two major developments introduced this year: adaptive sampling for limit surface sampling and multi variate distributions. The document concludes with a description of the demo case (BWR-SBO) and a discussion of the results obtained.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Deep-learning methods for contrast enhancement and artifact reduction in cryo-electron tomography: a systematic analysis of the state of the art and proposed improvements

Cryo-electron tomography (cryo-ET) has emerged as the preferred technique for visualizing the organization of macromolecular complexes in situ and resolving their structures at subnanometre resolution [Tegunov et al. (2021)View full citation, Nat. Methods, 18, 186–193]. Despite improvements in data quality as a result of advances in detector technology, microscope stability and stage precision, the analysis and interpretation of tomograms remains challenging due to a low signal-to-noise ratio and reconstruction artifacts stemming from experimental constraints in specimen tilt during data collection resulting in a missing wedge in the Fourier space. Recently, self-supervised deep-learning methods have been proposed for contrast enhancement and reduction of resolution anisotropy in reconstructed tomograms. Here, we evaluate several state-of-the-art deep-learning methods which aim to improve the interpretability of cryo-ET reconstructions, with a focus on their performance on downstream tasks of template matching, sub­tomogram averaging and segmentation. We propose new training architectures and a loss function based on Fourier shell correlation that show improved performance over the standard U-Net with L1/L2 losses. We demonstrate our analysis on four diverse experimental datasets: purified 80S ribosomes, in situ Chlamydomonas reinhardtii, immature HIV-1 virus-like particles and INS-1E cells.

contrast enhancement↗
DOE OSTI2025

Vision Foundation Models in Remote Sensing: A survey

Artificial intelligence (AI) technologies have profoundly transformed the field of remote sensing (RS), revolutionizing data collection, processing, and analysis. Traditionally reliant on manual interpretation and task-specific models, RS research has been significantly enhanced by the advent of foundation models (FMs)—large-scale pretrained AI models capable of performing a wide array of tasks with unprecedented accuracy and efficiency. This article provides a comprehensive survey of FMs in the RS domain. We categorize these models based on their architectures, pretraining datasets, and methodologies. Through detailed performance comparisons, we highlight emerging trends and the significant advancements achieved by those FMs. Additionally, we discuss technical challenges, practical implications, and future research directions, addressing the need for high-quality data, computational resources, and improved model generalization. Our research also finds that pretraining methods, particularly self-supervised learning (SSL) techniques like contrastive learning (CL) and masked autoencoders (MAEs), remarkably enhance the performance and robustness of FMs. This survey aims to serve as a resource for researchers and practitioners by providing a panorama of advances and promising pathways for the continued development and application of FMs in RS.

data models↗
DOE OSTI2018

Rapidly Assessing the Quality of Targeted Proteomics Experiments through Monitoring Stable-Isotope Labeled Standards

Selected reaction monitoring (SRM), also known as multiple reaction monitoring (MRM), is a data acquisition technique used in targeted analysis of molecules, including targeted proteomic studies. It exploits the unique capability of triple-quadrupole (QQQ) mass spectrometers to monitor the predefined precursor and fragment ion pairs of specific molecules of interest throughout a liquid chromatography (LC) elution profile. Compared to shotgun proteomics, targeted proteomics using SRM has high selectivity, high sensitivity, and a wide linear dynamic range,(1-3) which makes it especially useful in the accurate and reproducible quantification of low-abundance proteins in highly complex biological samples. SRM has been widely used in the fields of biomarker discovery,(4-7) analysis of protein post-translational modifications(8) and characterization of biological protein networks.(4,9) In the recent years, multiple technical advances have greatly improved the throughput of SRM analyses, allowing for the quantification of hundreds of peptides in a single analysis.(6,7,10) For example, a single 800-plex SRM assay (e.g., 400 unlabeled and heavy-labeled peptide pairs and 2400 transitions with retention time scheduling) using ultrahigh-performance liquid chromatography (UHPLC) has been developed to quantify proteins in plasma.(11) Advanced labeling techniques utilizing in vitro proteome-assisted MRM for protein absolute quantification (iMPAQT) demonstrated the capability of SRM in genome-wide protein quantification of over 18?000 human proteins.(12) Moreover, the scan speed of QQQ instrumentation has been greatly improved in recent releases of commercial instrumentation. The newly developed TSQ Altis (released in 2017) can scan more than 600 transitions per second, which is 6 times more than a traditional QQQ scan speed of 100 transitions per second. The breadth of measurement enabled by these technological improvements to QQQ mass spectrometry has increased the feasibility and popularity of large scale targeted proteomics studies,(13) a major application of which will be in clinical studies in which up to hundreds of protein candidates need to be quantified in hundreds of clinical specimens.(14) Quantitative accuracy is a primary motivating factor for utilizing a targeted proteomics protocol. The precise and reproducible absolute quantification produced by SRM assays is essential to many clinical and laboratory experiments.(15,16) Because the abundance of an endogenous peptide is calculated from the measurement of the spiked-in reference standard, it is essential to assess the data quality of these references.(17) In early applications of targeted proteomics, when instrument speed greatly limited the number of transitions that could be monitored, much of this quality assessment was done manually. However, recent improvements in instrument performance and experimental design have enabled a dramatic increase in the number of target peptides and associated SRM transitions, which makes manual quality assessment an untenable and laborious task. A variety of computational tools assist in SRM experiment design and data analysis. The first task in creating an SRM experiment is the choice of proteins and representative peptides to monitor. Achieving a reliable protein abundance requires appropriately choosing peptides that have a strong signal and are free from interferences in the biological matrix. Numerous computational tools exist to facilitate assay design by identifying peptides and refining SRM transitions.(18-21) To help share these assays and eliminate time spent designing the same transitions at multiple institutions, community portals have begun to host well-designed and vetted assays.(22-24) Analyzing the experimental data requires significant computational effort to align files across replicates and experimental conditions, pick peaks and produce quantitative values, normalize data and perform statistical tests, etc.(25-27)

Gibbons, Bryson C.↗
DOE OSTI2023

Deep learning-enabled natural language processing to identify directional pharmacokinetic drug–drug interactions

Background. During drug development, it is essential to gather information about the change of clinical exposure of a drug (object) due to the pharmacokinetic (PK) drug-drug interactions (DDIs) with another drug (precipitant). While many natural language processing (NLP) methods for DDI have been published, most were designed to evaluate if (and what kind of) DDI relationships exist in the text, without identifying the direction of DDI (object vs. precipitant drug). Here we present a method for the automatic identification of the directionality of a PK DDI from literature or drug labels. Methods. We reannotated the Text Analysis Conference (TAC) DDI track 2019 corpus for identifying the direction of a PK DDI and evaluated the performance of a fine-tuned BioBERT model on this task by following the training and validation steps prespecified by TAC. Results. This initial attempt showed the model achieved an F-score of 0.82 in identifying sentences as containing PK DDI and an F-score of 0.97 in identifying object versus precipitant drugs in those sentences. Discussion and conclusion. Despite a growing list of NLP methods for DDI extraction, most of them use a common set of corpora to perform general purpose tasks (e.g., classifying a sentence into one of several fixed DDI categories). There is a lack of coordination between the drug development and biomedical informatics method development community to develop corpora and methods to perform specific tasks (e.g., extract clinical exposure changes due to PK DDI). We hope that our effort can encourage such a coordination so that more “fit for purpose” NLP methods could be developed and used to facilitate the drug development process.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2022

ORNL Neutron Cross Section Measurements of 90 Zr

Nuclear criticality modeling and simulations rely on the quality of the existing evaluated nuclear data libraries such as Evaluated Nuclear Data File (ENDF)/B, the Joint Evaluated Fission and Fusion (JEFF) nuclear data library, or the Japanese Evaluated Nuclear Data Library (JENDL). In some cases, the cross-section evaluations of those libraries were found to be deficient in describing criticality benchmarks accurately. More than two decades ago, the US Nuclear Criticality Safety Program (NCSP) established a Nuclear Data (ND) task which encompassed experiments and evaluations. In response to this, the Oak Ridge National Laboratory (ORNL) formed a Nuclear Criticality and Data group which performed ND experiments, data analysis, and evaluations to produce ENDF files for the ND libraries as identified in the NCSP Five-Year Plan. Before being submitted to the ENDF library, files were processed and tested for performance by running benchmark calculations. This procedure was centralized in the ORNL group and is now often referred to as the ND pipeline. NCSP collaborates with the Joint Research Center (JRC) of the European Commission in Geel, Belgium, to perform high-resolution neutron-induced cross section measurements at the Geel Linear Accelerator (GELINA). The objective is to address emerging ND problems in criticality calculations. Difficulties with ND include insufficient neutron energy range, missing covariances, and previously unrecognized inaccuracies with experiments. New neutron total and capture cross sections of 90 Zr in the neutron energy range from 100 eV to several hundred keV were recently performed. These measured data will be used, together with existing high-resolution transmission data from a metallic 90 Zr sample, to improve representation of the cross sections.

97 MATHEMATICS AND COMPUTING↗