SEARCH · Engineering Papers
Results for “label shift”
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.
Computer-aided Abnormality Detection in Chest Radiographs in a Clinical Setting via Domain-adaptation
Deep learning (DL) models are being deployed at medical centers to aid radiologists for diagnosis of lung conditions from chest radiographs. Such models are often trained on a large volume of publicly available labeled radiographs. These pre-trained DL models’ ability to generalize in clinical settings is poor because of the changes in data distributions between publicly available and privately held radiographs. In chest radiographs, the heterogeneity in distributions arises from the diverse conditions in X-ray equipment and their configurations used for generating the images. In the machine learning community, the challenges posed by the heterogeneity in the data generation source is known as domain shift, which is a mode shift in the generative model. In this work, we introduce a domain-shift detection and removal method to overcome this problem. Our experimental results show the proposed method’s effectiveness in deploying a pre-trained DL model for abnormality detection in chest radiographs in a clinical setting.
OpenPATH - Leveraging Technology to Measure Travel Behavior
Shifting transportation to more sustainable modes is a key piece of the decarbonization puzzle. However, mobility behavior and travel patterns are difficult to influence because they are difficult to measure. OpenPATH provides a tool to capture longitudinal behaviors through a smartphone application. Agencies interested in gathering data about a population's travel behavior can set up a deployment of the app customized to the needs of their community. Partners can choose between simple mode and purpose labels or surveys for each trip to balance the level of user engagement with the associated burden. The labels, trip surveys, and an initial demographic survey can all be tailored to the specific context of the deployment. The OpenPATH tool is unique in its open-source nature, ability to gather detailed longitudinal travel data, and design allowing direct engagement with travelers. A valuable technological advancement, this tool enables partners to measure the way changes in the transportation landscape impact their community. The suite of tools includes both public and administrator dashboards. The public dashboard supports continuous data analysis through charts presenting trip information updated daily. The administrator dashboard displays geospatial data and supports data export. Example applications have included e-bike programs; gathering valuable metrics on increased access to opportunities and reduction in VMT, and studies aimed at understanding existing mobility behavior to see where advancements such as electric vehicles could fit into these habits. OpenPATH collects travel data in association with an initial demographic survey, enabling detailed insight into the behavior patterns or impact of a certain program on different populations.
Luminescence of Lanthanide-Dimethyl Sulfoxide Compound Solutions
Dimethyl sulfoxide (DMSO) has the ability to penetrate living tissues without causing significant damage. Of foremost importance to our understanding of the possible functions of DMSO in biological systems is its ability to replace some of the water molecules associated with the cellular constituents, or to affect the structure of the omnipresent water. Luminescence probes have been widely used for biological studies such as labeling, imaging and detection. Luminescence probes formed in DMSO may find new applications. Here, luminescence compounds formed by refluxing lanthanide nitrates of Ce, La, Tb, Yb, Nd, Gd and Eu in DMSO are reported and their luminescence properties investigated. Based on their luminescence spectral properties, the compounds can be classified into four classes. For compounds-I with Yb, Ce, and La, the excitation and emission spectra are very broad and their excitation or emission peaks are shifted to longer wavelengths when the monitored emission or excitation wavelength is longer . For compounds-II with Gd and Nd, both the excitation and emission spectra are very broad but their emission wavelengths change little at different excitation wavelengths. For Tb-DMSO as compound-III, both the typical emissions from the f - f transitions of Tb3+ and a broad emission at 445 nm are observed. At low temperatures of reaction, the f - f emissions are dominant, while at high temperatures such as 180 oC of reaction, the broad emission at 445 nm is dominant. For compound-IV with Eu-DMSO compounds, the dominant emissions are from the f - f transitions of Eu3+ and only a weak broad emission is observed, which is likely from the d - f transition of Eu2+ rather than from the metal to ligand charge transfer states.
Toward photostable multiplex analyte detection on a single mode planar optical waveguide
We have developed a waveguide-based optical biosensor for the sensitive and specific detection of biomarkers associated with disease. Our technology combines the superior optical properties of single-mode planar waveguides, the robust nature of functionalized self-assembled monolayer sensing films and the specificity of fluorescence sandwich immunoassays to detect biomarkers in complex biological samples such as serum, urine and sputum. We have previously reported the adaptation of our technology to the detection of biomarkers associated with breast cancer and anthrax. However, these approaches primarily used phospholipid bilayers as the functional film and organic dyes (ex: AlexaFluors) as the fluorescence reporter. Organic dyes are easily photodegraded and are not amenable to multiplexing because of their narrow Stokes' shift. Here we have developed strategies for conjugation of the detector antibodies with quantum dots for use in a multiplex detection platform. We have previously evaluated dihydroxylipoic acid quantum dots for the detection of a breast cancer biomarker. In this manuscript, we investigate the detection of the Bacillus anthracis protective antigen using antibodies conjugated with polymer-coated quantum dots. Kinetics of binding on the waveguide-based biosensor is reported. We compare the sensitivity of quantum dot labeled antibodies to those labeled with AlexaFluor and demonstrate the photostability of the former in our assay platform. In addition, we compare sulfydryl labeling of the antibody in the hinge region to that of nonspecific amine labeling. This is but the first step in developing a multiplex assay for such biomarkers on our waveguide platform.
Resource-Adaptive Federated Text Generation with Differential Privacy
In cross-silo federated learning (FL), sensitive text datasets remain confined to local organizations due to privacy regulations, making repeated training for each downstream task both communication-intensive and privacy-demanding. A promising alternative is to generate differentially private (DP) synthetic datasets that approximate the global distribution and can be reused across tasks. However, pretrained large language models (LLMs) often fail under domain shift, and federated finetuning is hindered by computational heterogeneity: only resource-rich clients can update the model, while weaker clients are excluded, amplifying data skew and the adverse effects of DP noise. We propose a flexible participation framework that adapts to client capacities. Strong clients perform DP federated finetuning, while weak clients contribute through a lightweight DP voting mechanism that refines synthetic text. To ensure the synthetic data mirrors the global dataset, we apply control codes (e.g., labels, topics, metadata) that represent each client’s data proportions and constrain voting to semantically coherent subsets. This two-phase approach requires only a single round of communication for weak clients and integrates contributions from all participants. Experiments show that our framework improves distribution alignment and downstream robustness under DP and heterogeneity.
Unraveling the Dynamic Network in the Reactions of an Alkyl Aryl Ether Catalyzed by Ni/γ-Al 2 O 3 in 2-Propanol
The reductive cleavage of aryl ether linkages is a key step in the disassembly of lignin to its monolignol components, where selectivity is determined by the kinetics of multiple parallel and consecutive liquid-phase reactions. Triphasic hydrogenolysis of 13 C-labeled benzyl phenyl ether (BPE, a model compound with the same C-O bond energy as the major β-O-4 linkage in lignin), catalyzed by Ni/γ-Al 2 O 3 , was observed directly at elevated temperatures (150 - 175 °C) and pressures (79-89 bar) using operando magic-angle spinning NMR spectroscopy. Liquid-vapor partitioning in the NMR rotor was quantified using the 13 C NMR resonances for the 2-propanol solvent, whose chemical shifts report on the internal reactor temperature. At 170 °C, BPE is converted to toluene and phenol with k 1 = 0.17 s -1 g cat -1 and an apparent activation barrier of (80 ± 8) kJ mol -1 . Subsequent phenol hydrogenation occurs much more slowly (k 2 = 0.0052 s -1 g cat -1 at 170-175 °C), such that cyclohexanol formation is significant only at higher temperatures. Toluene is stable under these reaction conditions, but its methyl group undergoes facile H/D exchange (k 3 = 0.046 s -1 g cat -1 at 175 °C). While the source of the reducing equivalents for both hydrogenolysis and hydrogenation is exclusively H 2 /D 2(g) rather than the alcohol solvent at these temperatures, the initial isotopic composition of adsorbed H/D on the catalyst surface is principally determined by the solvent isotopic composition (2-PrOH/D). All reactions are preceded by a pronounced induction period associated with catalyst activation. In air, Ni nanoparticles are passivated by a surface oxide monolayer, whose removal under H 2 proceeds with an apparent activation barrier of (72 ± 13) kJ mol -1 . Finally ,the operando NMR spectra provide molecularly-specific, time-resolved information about the multiple simultaneous and sequential processes as they occur at the solid-liquid interface.
Unlocking plant-microbial interactions in deep Mollisols in the Midwestern US: Linking depth gradients in roots, microbial activity, and soil carbon in agroecosystems
Deep-rooted plants may build soil carbon (C) stocks, but most research has focused on shallow soils, leaving gaps in our understanding of how shifts in the balance between decomposition and C inputs drive soil C accumulation with depth. Thus, our objectives were to: (1) link depth gradients in root biomass with microbial activity and soil C stocks down to 1 m, and (2) examine the potential of simple C inputs to prime soil C across depths. To this end, we dug 5 quantitative soil pits in Argiudolls under mature perennial miscanthus plots in the SoyFACE Farm (Champaign-Urbana, IL). We added 13 C labeled glucose to our soils to determine the fate of simple C inputs with depth. We found that fine root biomass, total soil C, mineral-associated organic C (MAOC), particulate organic C (POC), and microbial activity (as measured by potential enzyme activity) declined with depth. POC declined more rapidly than MAOC, resulting in an increase in the ratio of MAOC-to-POC. Root biomass, enzyme activity (either acid phosphatase or n-acetyl-glucosaminadase) activity, and microbial respiration explained 74% and 38% of the variability in soil total C and MAOC, respectively, while POC was dependent on root biomass and microbial respiration (47%). Although the incorporation of simple 13 C inputs into MAOC was similar across depths, these inputs led to greater net MAOC losses in shallow soils than in deeper soils between 50 and 100 cm. The divergent impact of simple C inputs across depths may suggest that MAOC in shallow soils is more susceptible to priming losses, while C inputs into deep soils may instead be more persistent. Collectively, our results suggest that depth gradients in soil C stocks represents a balance between inputs, decomposition, and microbial necromass production and that increases in root C inputs by deep-rooted plants may have the potential to build stable MAOC.
Photonic crystal nanobeam biosensors based on porous silicon
Photonic crystal (PhC) nanobeams (NB) patterned on porous silicon (PSi) waveguide substrates are demonstrated for the specific, label-free detection of oligonucleotides. These photonic structures combine the large active sensing area intrinsic to PSi sensors with the high-quality (Q) factor and low-mode volume characteristic of compact resonant silicon-on-insulator (SOI) PhC NB devices. The PSi PhC NB can achieve a Q-factor near 9,000 and has an approximately 40-fold increased active sensing area for molecular attachment, compared to traditional SOI PhC NB sensors. The PSi PhC NB exhibits a resonance shift that is more than one order of magnitude larger than that of a similarly designed SOI PhC NB for the detection of small chemical molecules and 16-base peptide nucleic acids. Furthermore the design and fabrication of PSi PhC NB sensors are compatible with CMOS processing, sensor arrays, and integration with lab-on-chip systems.
Aging Impairs Myocardial Fatty Acid and Ketone Oxidation and Modifies Cardiac Functional and Metabolic Responses to Insulin in Mice
Aging presumably initiates shifts in substrate oxidation mediated in part by changes in insulin sensitivity. Similar shifts occur with cardiac hypertrophy and may contribute to contractile dysfunction. We tested the hypothesis that aging modifies substrate utilization and alters insulin sensitivity in mouse heart when provided multiple substrates. In vivo cardiac function was measured with microtipped pressure transducers in the left ventricle from control (4–6 mo) and aged (22–24 mo) mice. Cardiac function was also measured in isolated working hearts along with substrate and anaplerotic fractional contributions to the citric acid cycle (CAC) by using perfusate containing 13C-labeled free fatty acids (FFA), acetoacetate, lactate, and unlabeled glucose. Stroke volume and cardiac output were diminished in aged mice in vivo, but pressure development was preserved. Systolic and diastolic functions were maintained in aged isolated hearts. Insulin prompted an increase in systolic function in aged hearts, resulting in an increase in cardiac efficiency. FFA and ketone flux were present but were markedly impaired in aged hearts. These changes in myocardial substrate utilization corresponded to alterations in circulating lipids, thyroid hormone, and reductions in protein expression for peroxisome proliferator-activated receptor (PPAR)α and pyruvate dehydrogenase kinase (PDK)4. Insulin further suppressed FFA oxidation in the aged. Insulin stimulation of anaplerosis in control hearts was absent in the aged. The aged heart shows metabolic plasticity by accessing multiple substrates to maintain function. However, fatty acid oxidation capacity is limited. Impaired insulin-stimulated anaplerosis may contribute to elevated cardiac efficiency, but may also limit response to acute stress through depletion of CAC intermediates.
AUTOMATIC UNSUPERVISED CLASSIFICATION OF ALL SLOAN DIGITAL SKY SURVEY DATA RELEASE 7 GALAXY SPECTRA
Using the k-means cluster analysis algorithm, we carry out an unsupervised classification of all galaxy spectra in the seventh and final Sloan Digital Sky Survey data release (SDSS/DR7). Except for the shift to rest-frame wavelengths and the normalization to the g-band flux, no manipulation is applied to the original spectra. The algorithm guarantees that galaxies with similar spectra belong to the same class. We find that 99% of the galaxies can be assigned to only 17 major classes, with 11 additional minor classes including the remaining 1%. The classification is not unique since many galaxies appear in between classes; however, our rendering of the algorithm overcomes this weakness with a tool to identify borderline galaxies. Each class is characterized by a template spectrum, which is the average of all the spectra of the galaxies in the class. These low-noise template spectra vary smoothly and continuously along a sequence labeled from 0 to 27, from the reddest class to the bluest class. Our Automatic Spectroscopic K-means-based (ASK) classification separates galaxies in colors, with classes characteristic of the red sequence, the blue cloud, as well as the green valley. When red sequence galaxies and green valley galaxies present emission lines, they are characteristic of active galactic nucleus activity. Blue galaxy classes have emission lines corresponding to star formation regions. We find the expected correlation between spectroscopic class and Hubble type, but this relationship exhibits a high intrinsic scatter. Several potential uses of the ASK classification are identified and sketched, including fast determination of physical properties by interpolation, classes as templates in redshift determinations, and target selection in follow-up works (we find classes of Seyfert galaxies, green valley galaxies, as well as a significant number of outliers). The ASK classification is publicly accessible through various Web sites.
The enduring role of contracts for difference in risk management and market creation for renewables
Governments procure renewables through a variety of mechanisms. Contracts for difference (CfDs) have been used for more than 50% of the global offshore wind supply. The payments awarded through CfDs are sometimes labelled subsidies, suggesting that they support uneconomic activity. Here, in this study, we argue that the primary role of CfDs is rather risk management by creating a market for electricity supply at stable long-term prices. Similar to its use in other sectors of the economy, this contract type transforms a variable to a fixed price to reallocate volatility risks. Such long-term contracts are often necessary for renewables financing due to limited hedging options in existing markets. Our perspective could imply a shift in perception towards CfDs as a fundamental and lasting market feature. We hope to stimulate a timely discussion about the impact of greater CfD diffusion on electricity market mechanisms, risk allocation and the potential for combining fragmented streams of energy finance, market and policy research.
Changes in Soil Carbon Dynamics in Response to Long-Term Soil Warming - Integration Across Scales from Cells to Ecosystems
This project produced a series of papers documenting the effects of soil warming on fluxes of carbon dioxide between soils and the atmosphere, and the temporal changes in structure and function of soil microbial community. A diversity of techniques was used in the study including: pyrolysis–GC/MS to analyze soil chemistry; isotopically–labeled organic compounds to estimate microbial carbon use efficiency; and Illumina sequencing of the 16S ribosomal RNA gene to study bacterial community composition. We found that soil warming resulted in a four–phase pattern of soil organic matter decay and carbon dioxide fluxes to the atmosphere, with phases of substantial soil carbon loss alternating with phases of no detectable loss. Several factors combined to affect the timing, magnitude, and thermal acclimation of soil carbon loss. These include depletion of microbially accessible carbon pools, reductions in microbial biomass, a shift in microbial carbon use efficiency, and changes in microbial community composition. Our results support projections of a long–term, self–reinforcing carbon feedback from mid–latitude forests to the climate system as the world warms.
Interactions of Oxide Surfaces with Water Revealed with Solid-State NMR Spectroscopy
Hydrous materials are ubiquitous in the natural environment and efforts have previously been made to investigate the structures and dynamics of hydrated surfaces for their key roles in various chemical and physical applications, with the help of theoretical modeling and microscopy techniques. However, an overall atomic-scale understanding of the water–solid interface, including the effect of water on surface ions, is still lacking. Herein, we employ ceria nanorods with different amounts of water as an example and demonstrate a new approach to explore the water–surface interactions by using solid-state NMR in combination with density functional theory. NMR shifts and relaxation time analysis provide detailed information on the local structure of oxygen ions and the nature of water motion on the surface: the amount of molecularly adsorbed water decreases rapidly with increasing temperature (from room temperature to 150 °C), whereas hydroxyl groups are stable up to 150 °C, and dynamic water molecules are found to instantaneously coordinate to the surface oxygen ions. Additionally, the applicability of dynamic nuclear polarization for selective detection of surface oxygen species is also compared to conventional NMR with surface selective isotopic-labeling: the optimal method depends on the feasibility of enrichment and the concentration of protons in the sample. These results provide new insight into the interfacial structure of hydrated oxide nanostructures, which is important to improve performance for various applications.
NMR relaxation of the orientation of single segments in semiflexible dendrimers
We study the orientational properties of labeled segments in semiflexible dendrimers making use of the viscoelastic approach of Dolgushev and Blumen [J. Chem. Phys. 131, 044905 (2009)]. We focus on the segmental orientational autocorrelation functions (ACFs), which are fundamental for the frequency-dependent spin-lattice relaxation times T{sub 1}(ω). We show that semiflexibility leads to an increase of the contribution of large-scale motions to the ACF. This fact influences the position of the maxima of the [1/T{sub 1}]-functions. Thus, going from outer to inner segments, the maxima shift to lower frequencies. Remarkably, this feature is not obtained in the classical bead-spring model of flexible dendrimers, although many experiments on dendrimers manifest such a behavior.
Simple eigenvalue-self-consistent $\overline{Δ}GW_0$
We show that a rigid scissors-like GW self-consistency approach, labeled here $\overline{Δ}GW_0$, can be trivially implemented at zero additional cost for large scale one-shot $G_0W_0$ calculations. The method significantly improves one-shot $G_0W_0$ and for large systems is very accurate. $\overline{Δ}GW_0$ is similar in spirit to ev$GW_0$ where the self-consistency is only applied on the eigenvalues entering Green's function, while both W and the eigenvectors of Green's function are held fixed. $\overline{Δ}GW_0$ further assumes that the shift of the eigenvalues is rigid scissors-like so that all occupied states are shifted by the same amount and analogously for all the unoccupied states. We show that this results in a trivial modification of the time-dependent $G_0W_0$ self-energy, enabling an a posteriori self-consistency cycle. The method is applicable for our recent stochastic-$GW$ approach, thereby enabling self-consistent calculations for giant systems with thousands of electrons. The accuracy of $\overline{Δ}GW_0$ increases with the system size. For molecules, it is up to 0.4-0.5 eV away from coupled-cluster single double triple (CCSD(T)), but for tetracene and hexacene, it matches the ionization energies from both CCSD(T) and ev$GW_0$ to better than 0.05 eV. For solids, as exemplified here by periodic supercells of semiconductors and insulators with 6192 valence electrons, the method matches ev$GW_0$ quite well and both methods are in good agreement with the experiment.
HydroDCM: Hydrological Domain-Conditioned Modulation for Cross-Reservoir Inflow Prediction
Deep learning models have shown promise in reservoir inflow prediction, yet their performance often deteriorates when applied to different reservoirs due to distributional differences, referred to as the domain shift problem. Domain generalization (DG) solutions aim to address this issue by extracting domain-invariant representations that mitigate errors in unseen domains. However, in hydrological settings, each reservoir exhibits unique inflow patterns, while some metadata beyond observations like spatial information exerts indirect but significant influence. This mismatch limits the applicability of conventional DG techniques to many-domain hydrological systems. To overcome these challenges, we propose HydroDCM, a scalable DG framework for cross-reservoir inflow forecasting. Spatial metadata of reservoirs is used to construct pseudo-domain labels that guide adversarial learning of invariant temporal features. During inference, HydroDCM adapts these features through light-weight conditioning layers informed by the target reservoir’s metadata, reconciling DG’s invariance with location-specific adaptation. Experiment results on 30 real-world reservoirs in the Upper Colorado River Basin demonstrate that our method substantially outperforms state-of-the-art DG baselines under many-domain conditions and remains computationally efficient.
Microbial Community Shifts Reflect Losses of Native Soil Carbon with Pyrogenic and Fresh Organic Matter Additions and Are Greatest in Low-Carbon Soils
ABSTRACT Soil organic carbon (SOC) plays an important role in regulating global climate change, carbon and nutrient cycling in soils, and soil moisture. Organic matter (OM) additions to soils can affect the rate at which SOC is mineralized by microbes, with potentially important effects on SOC stocks. Understanding how pyrogenic organic matter (PyOM) affects the cycling of native SOC (nSOC) and the soil microbes responsible for these effects is important for fire-affected ecosystems as well as for biochar-amended systems. We used an incubation trial with five different soils from National Ecological Observatory Network sites across the United States and 13 C-labeled 350°C corn stover PyOM and fresh corn stover OM to trace nSOC-derived CO 2 emissions with and without PyOM and OM amendments. We used high-throughput sequencing of rRNA genes to characterize bacterial, archaeal, and fungal communities and their responses to PyOM and OM in soils that were previously stored at −80°C. We found that the effects of amendments on nSOC-derived CO 2 reflected the unamended soil C status, where relative increases in C mineralization were greatest in low-C soils. OM additions produced much greater effects on nSOC-CO 2 emissions than PyOM additions. Furthermore, the magnitude of the microbial community composition change mirrored the magnitude of increases in nSOC-CO 2 , indicating that a specific subset of microbes was likely responsible for the observed changes in nSOC mineralization. However, PyOM responders differed across soils and did not necessarily reflect a common “charosphere.” Overall, this study suggests that soils that already have low SOC may be particularly vulnerable to short-term increases in SOC loss with OM or PyOM additions. IMPORTANCE Soil organic matter (SOM) has an important role in global climate change, carbon and nutrient cycling in soils, and soil moisture dynamics. Understanding the processes that affect SOM stocks is important for managing these functions. Recently, understanding how fire-affected organic matter (or “pyrogenic” organic matter [PyOM]) affects existing SOM stocks has become increasingly important, due to both changing fire regimes and interest in “biochar,” pyrogenic organic matter that is produced intentionally for carbon management or as an agricultural soil amendment. We found that soils with less SOM were more prone to increased losses with PyOM (and fresh organic matter) additions and that soil microbial communities changed more in soils that also had greater SOM losses with PyOM additions. This suggests that soils that already have low SOM content may be particularly vulnerable to short-term increases in SOM loss and that a subset of the soil microbial community is likely responsible for these effects.