Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “feature map”

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 415 records · Page 23
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

The benefits of CMB delensing

The effects of gravitational lensing of the cosmic microwave background (CMB) have been measured at high significance with existing data and will be measured even more precisely in future surveys. Reversing the effects of lensing on the observed CMB temperature and polarization maps provides a variety of benefits. Delensed CMB spectra have sharper acoustic peaks and more prominent damping tails, allowing for improved inferences of cosmological parameters that impact those features. Delensing reduces B-mode power, aiding the search for primordial gravitational waves and allowing for lower variance reconstruction of lensing and other sources of secondary CMB anisotropies. Lensing-induced power spectrum covariances are reduced by delensing, simplifying analyses and improving constraints on primordial non-Gaussianities. Biases that result from incorrectly modeling nonlinear and baryonic feedback effects on the lensing power spectrum are mitigated by delensing. All of these benefits are possible without any changes to experimental or survey design. Here, we develop a self-consistent, iterative, all-orders treatment of CMB delensing on the curved sky and demonstrate the impact that delensing will have with future surveys.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2021

Improving deep learning-based protein distance prediction in CASP14

Abstract Motivation Accurate prediction of residue–residue distances is important for protein structure prediction. We developed several protein distance predictors based on a deep learning distance prediction method and blindly tested them in the 14th Critical Assessment of Protein Structure Prediction (CASP14). The prediction method uses deep residual neural networks with the channel-wise attention mechanism to classify the distance between every two residues into multiple distance intervals. The input features for the deep learning method include co-evolutionary features as well as other sequence-based features derived from multiple sequence alignments (MSAs). Three alignment methods are used with multiple protein sequence/profile databases to generate MSAs for input feature generation. Based on different configurations and training strategies of the deep learning method, five MULTICOM distance predictors were created to participate in the CASP14 experiment. Results Benchmarked on 37 hard CASP14 domains, the best performing MULTICOM predictor is ranked 5th out of 30 automated CASP14 distance prediction servers in terms of precision of top L/5 long-range contact predictions [i.e. classifying distances between two residues into two categories: in contact (<8 Angstrom) and not in contact otherwise] and performs better than the best CASP13 distance prediction method. The best performing MULTICOM predictor is also ranked 6th among automated server predictors in classifying inter-residue distances into 10 distance intervals defined by CASP14 according to the precision of distance classification. The results show that the quality and depth of MSAs depend on alignment methods and sequence databases and have a significant impact on the accuracy of distance prediction. Using larger training datasets and multiple complementary features improves prediction accuracy. However, the number of effective sequences in MSAs is only a weak indicator of the quality of MSAs and the accuracy of predicted distance maps. In contrast, there is a strong correlation between the accuracy of contact/distance predictions and the average probability of the predicted contacts, which can therefore be more effectively used to estimate the confidence of distance predictions and select predicted distance maps. Availability and implementation The software package, source code and data of DeepDist2 are freely available at https://github.com/multicom-toolbox/deepdist and https://zenodo.org/record/4712084#.YIIM13VKhQM. Supplementary information Supplementary data are available at Bioinformatics online.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2021

Development of Automated Pipeline for Time-Resolved Link-Wise Vehicular Energy Consumption in the Chattanooga, TN Road Network

The Department of Energy (DOE) has shown strong interest in detecting energy inefficiencies in regional road networks, so as to derive energy consumed at a high spatial temporal resolution. We have developed a workflow to automate the estimation of time-resolved vehicular energy consumption over each link in a road network of interest. The road network used in the current work is centered around the city of Chattanooga, Tennessee and its bordering regions. Utilizing the most mature road network for the Chattanooga, TN region, vehicle speed & count data from TomTom in conjunction with machine learning methods, we have developed an automated pipeline to estimate energy consumption for every link in the network. The first step in the pipeline is ingesting vehicle probe counts and speed estimates from TomTom API. In the next step, the probe counts, speed profiles and other exogenous data (i.e. road types, weather data, ground-truth volume counts and more) were used as input to a supervised learning algorithm to estimate the number of vehicles throughout the entire region for each road segment. These volume estimates were then mapped to a unified road network that contained additional important information such as percentage change in gradient across a link, number of lanes and link lengths that are features in pre-trained single vehicle energy-consumption models available with the RouteE software developed at NREL. The per vehicle energy consumption on each road link predicted using appropriate RouteE vehicular models were multiplied by the volume estimate for the corresponding link over a given time period to predict energy consumed per link for the time interval of interest. Currently, work is underway to improve both the RouteE per vehicle energy estimate and the volume estimates derived from TomTom probe counts. We have also explored the correlation of the link-wise energy estimates with the features of the pre-trained RouteE machine learning model in order to gain insight into what factors contribute most to the link-wise energy consumption.

ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATION,↗
DOE OSTI2025

WELLS Interactive Application

The Wellbore Exploration and Location Logistic System (WELLS) Interactive Application is an interactive tool to enable easy exploration and visualization of the living national wellbore database (WELLS Database (https://edx.netl.doe.gov/dataset/wells_database)). The tool and underlying database were created and are maintained by the National Energy Technology Laboratory (NETL), providing visualization of the more than six million public wellbore records from more than 65 authoritative state, federal, and tribal resources. The WELLS Interactive Application serves up wellbore data from oil, gas, underground injection, research, geothermal, geotechnical, groundwater, and other types of wells in a single, standardized, unified system. In addition to the surface location of these wells, the underlying database combines select key attributes for features such as well age, depth, and operating status. The system also provides users with references back to the original sources used in this unified platform. The underlying data can be accessed through the WELLS Database: https://edx.netl.doe.gov/dataset/wells_database Additional Information: The WELLS Interactive Application (formerly titled CO2-Locate) enables visualization and access to the public wellbore records through an intuitive web-based mapping tool. The WELLS Interactive Application was designed to help users visualize, query, analyze, and download wellbore records. Public wellbore points are included as a layer in the Map page, called Public Wells. Additionally, a multivariate hexagon grid summarizing well density from proprietary well data, called Well Density, is included to identify data gaps between the public and proprietary well data. Filtering functionalities in the tool allow these two layers to be spatially filtered by state, county, or basin as well as by status, type, true vertical depth, and spud year. The WELLS Interactive Application also contains a Near Me tool can be used to search and explore wellbore data within a user-defined distance of a specified location on the map, which can also be downloaded. The Query tool allows users to query the selected or filtered wells in the Public Wells layer and export the data. For additional information on these tool functionalities, see the help documentation on the About page of the tool. Notes for Consideration: The Well Density layer provided in this application is derived from proprietary wellbore data, the records of which do not always contain values for key features (status, type, true vertical depth, or spud year). Therefore, data might not be available when layers are queried for all filter combinations. Additionally, visualizing layers and applying filters may take additional time to load (i.e., draw on the map) due to the large size of the data.

ccs↗
DOE OSTI2022

A detailed map of Higgs boson interactions by the ATLAS experiment ten years after the discovery

The standard model of particle physics describes the known fundamental particles and forces that make up our Universe, with the exception of gravity. One of the central features of the standard model is a field that permeates all of space and interacts with fundamental particles. The quantum excitation of this field, known as the Higgs field, manifests itself as the Higgs boson, the only fundamental particle with no spin. In 2012, a particle with properties consistent with the Higgs boson of the standard model was observed by the ATLAS and CMS experiments at the Large Hadron Collider at CERN. Since then, more than 30 times as many Higgs bosons have been recorded by the ATLAS experiment, enabling much more precise measurements and new tests of the theory. Here, on the basis of this larger dataset, we combine an unprecedented number of production and decay processes of the Higgs boson to scrutinize its interactions with elementary particles. Interactions with gluons, photons, and W and Z bosons—the carriers of the strong, electromagnetic and weak forces—are studied in detail. Interactions with three third-generation matter particles (bottom (b) and top (t) quarks, and tau leptons (τ)) are well measured and indications of interactions with a second-generation particle (muons, μ) are emerging. These tests reveal that the Higgs boson discovered ten years ago is remarkably consistent with the predictions of the theory and provide stringent constraints on many models of new phenomena beyond the standard model.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2022

Mapping periplasmic binding protein oligosaccharide recognition with neutron crystallography

Numerous studies have shown how periplasmic binding proteins (PBPs) bind substrates with exquisite specificity, even distinguishing between sugar epimers and anomers, or structurally similar ions. Yet, marked substrate promiscuity is also a feature encoded in some PBPs. Except for three sub-Ångström crystal structures, there are no reports of hydrogen atom positions in the remaining (> 1000) PBP structures. The previous X-ray crystal structure of the maltodextrin periplasmic-binding protein from Thermotoga maritima (tmMBP) complexed with oligosaccharide showed a large network of interconnected water molecules stretching from one end of the substrate binding pocket to the other. These water molecules are positioned to form multiple hydrogen bonds, as well as forming interactions between the protein and substrate. Here we present the neutron crystal structure of tmMBP to a resolution of 2.1 Å. This is the first neutron crystal structure from the PBP superfamily and here we unambiguously identify the nature and orientation of the hydrogen bonding and water-mediated interactions involved in stabilizing a tetrasaccharide in the binding site. More broadly, these results demonstrate the conserved intricate mechanisms that underlie substrate-specificity and affinity in PBPs.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2020

Planck intermediate results

We present the NPIPE processing pipeline, which produces calibrated frequency maps in temperature and polarization from data from the Planck Low Frequency Instrument (LFI) and High Frequency Instrument (HFI) using high-performance computers. NPIPE represents a natural evolution of previous Planck analysis efforts, and combines some of the most powerful features of the separate LFI and HFI analysis pipelines. For example, following the LFI 2018 processing procedure, NPIPE uses foreground polarization priors during the calibration stage in order to break scanning-induced degeneracies. Similarly, NPIPE employs the HFI 2018 time-domain processing methodology to correct for bandpass mismatch at all frequencies. In addition, NPIPE introduces several improvements, including, but not limited to: inclusion of the 8% of data collected during repointing manoeuvres; smoothing of the LFI reference load data streams; in-flight estimation of detector polarization parameters; and construction of maximally independent detector-set split maps. For component-separation purposes, important improvements include: maps that retain the CMB Solar dipole, allowing for high-precision relative calibration in higher-level analyses; well-defined single-detector maps, allowing for robust CO extraction; and HFI temperature maps between 217 and 857 GHz that are binned into 0'.9 pixels (Nside = 4096), ensuring that the full angular information in the data is represented in the maps even at the highest Planck resolutions. The net effect of these improvements is lower levels of noise and systematics in both frequency and component maps at essentially all angular scales, as well as notably improved internal consistency between the various frequency channels. Based on the NPIPE maps, we present the first estimate of the Solar dipole determined through component separation across all nine Planck frequencies. The amplitude is (3366.6 ± 2.7) μK, consistent with, albeit slightly higher than, earlier estimates. From the large-scale polarization data, we derive an updated estimate of the optical depth of reionization of τ = 0.051 ± 0.006, which appears robust with respect to data and sky cuts. There are 600 complete signal, noise and systematics simulations of the full-frequency and detector-set maps. As a Planck first, these simulations include full time-domain processing of the beam-convolved CMB anisotropies. The release of NPIPE maps and simulations is accompanied with a complete suite of raw and processed time-ordered data and the software, scripts, auxiliary data, and parameter files needed to improve further on the analysis and to run matching simulations.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2021

Connectivity‐based parcellation of normal and anatomically distorted human cerebral cortex

Abstract For over a century, neuroscientists have been working toward parcellating the human cortex into distinct neurobiological regions. Modern technologies offer many parcellation methods for healthy cortices acquired through magnetic resonance imaging. However, these methods are suboptimal for personalized neurosurgical application given that pathology and resection distort the cerebrum. We sought to overcome this problem by developing a novel connectivity‐based parcellation approach that can be applied at the single‐subject level. Utilizing normative diffusion data, we first developed a machine‐learning (ML) classifier to learn the typical structural connectivity patterns of healthy subjects. Specifically, the Glasser HCP atlas was utilized as a prior to calculate the streamline connectivity between each voxel and each parcel of the atlas. Using the resultant feature vector, we determined the parcel identity of each voxel in neurosurgical patients ( n = 40) and thereby iteratively adjusted the prior. This approach enabled us to create patient‐specific maps independent of brain shape and pathological distortion. The supervised ML classifier re‐parcellated an average of 2.65% of cortical voxels across a healthy dataset ( n = 178) and an average of 5.5% in neurosurgical patients. Our patient dataset consisted of subjects with supratentorial infiltrating gliomas operated on by the senior author who then assessed the validity and practical utility of the re‐parcellated diffusion data. We demonstrate a rapid and effective ML parcellation approach to parcellation of the human cortex during anatomical distortion. Our approach overcomes limitations of indiscriminately applying atlas‐based registration from healthy subjects by employing a voxel‐wise connectivity approach based on individual data.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2024

Parallel quantum computing simulations via quantum accelerator platform virtualization

Quantum circuit execution is a central task in quantum computation. Due to inherent quantum-mechanical constraints, quantum computing workflows often involve a considerable number of independent measurements over a large set of slightly different quantum circuits. Here we discuss a simple model for parallelizing such quantum circuit executions that is based on introducing a large array of virtual quantum processing units (mapped to HPC nodes in our case) as a parallel quantum computing platform. Implemented within the XACC framework, the model can readily take advantage of its backend-agnostic features, enabling parallel quantum computing/simulation over any target backend supported by XACC. We illustrate the performance of this approach by demonstrating strong scaling in two pertinent domain science problems, namely in computing the gradients for the multi-contracted variational quantum eigensolver and in data-driven quantum circuit learning, where we vary the number of qubits and the number of circuit layers. Here, the latter simulation leverages the cuQuantum library to run efficiently on GPU-accelerated HPC platforms.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Automating the detection of hydrological barriers and fragmentation in wetlands using deep learning and InSAR

The loss of hydrological connectivity and fragmentation of natural wetlands is a widespread driver of wetland degradation. Understanding where and how natural connectivity is impaired is essential for managing, protecting and remediating these ecosystems. Wetland Interferometric Synthetic Aperture Radar (Wetland InSAR) can provide information on surface flow orientation in wetlands at a high spatial resolution, which can be used for barrier detection. However, the broad application of this approach is constrained by the labour-intensive manual delineation of barriers based on mapped water levels. This study presents the first deep learning-based methodology for the automated detection of hydrological barriers. We trained a deep convolutional network to segment edge features of hydrological barriers in 25 image pairs captured by ALOS PALSAR-1 L-Band InSAR between 2006 and 2011. The training dataset consists of manually labelled and delineated barriers showing abrupt changes in water surface elevation and wrapped interferograms with high coherence. We tested this method across three wetland sites: the Everglades and southern Louisiana wetlands (United States) and the Cienaga de Zapata (Cuba). Across these sites, the convolutional network detected hydrological barriers with up to 84% accuracy. The model performed particularly well for linear hydrological barriers such as roads, dikes, and channels. Notably, some barriers impede flow only seasonally, appearing during low water levels and disappearing when water levels rise. Our automated approach to detecting and assessing wetland hydrologic connectivity can be applied more broadly to support the effective management of fragmented wetland ecosystems.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2020

Transport studies in a gate-tunable three-terminal Josephson junction

Josephson junctions with three or more superconducting leads have been predicted to exhibit topological effects in the presence of few conducting modes within the interstitial normal material. Such behavior, of relevance for topologically protected quantum bits, would lead to specific transport features measured between terminals, with topological phase transitions occurring as a function of phase and voltage bias. Although conventional, two-terminal Josephson junctions have been studied extensively, multiterminal devices have received relatively little attention to date. Motivated in part by the possibility to ultimately observe topological phenomena in multiterminal Josephson devices, as well as their potential for coupling gatemon qubits, here we describe the superconducting features of a top-gated mesoscopic three-terminal Josephson device. The device is based on an InAs two-dimensional electron gas proximitized by epitaxial aluminum. We map out the transport properties of the device as a function of bias currents, top gate voltage, and magnetic field. Finally, we find a very good agreement between the zero-field experimental phase diagram and a resistively and capacitively shunted junction computational model.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2025

Multiscale maps of Active Layer Depth for Teller site Mile Marker 27 and Kougarok Mile Marker 80, Seward Peninsula, AK

Remote sensing maps of active layer depth derived from Unmanned Areal System (UAS) data. The UAS datasets were stepwise scaled until matching the AVIRIS-NG (Airborne Visible / Infrared Imaging Spectrometer - Next Generation) and Sentinel-2 spatial resolutions. Using the field observed Active Layer Depth (ALD) measurement in combination with spectral and topographic predictors derivatives from DJI UAS imagery, we used a spatially explicit RF regression model to predict and map ALD across our study landscapes. This package includes maps for Next-Generation Ecosystem Experiment Arctic (NGEE Arctic)’s Teller Mile Marker (MM) 27, and Kougarok MM80 (aka Mile 80) watersheds. The field, map data, and metadata are provided as geoTIF and text (*.csv) formats. These datasets are provided in support of Hantson et al., 2024 (accepted) “Scaling Arctic landscape and permafrost features improves active layer depth modeling”

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

Dark Energy Survey Year 3 Results: Photometric Data Set for Cosmology

In this work, we describe the Dark Energy Survey (DES) photometric data set assembled from the first three years of science operations to support DES Year 3 cosmologic analyses, and provide usage notes aimed at the broad astrophysics community. Y3 GOLD improves on previous releases from DES, Y1 GOLD, and Data Release 1 (DES DR1), presenting an expanded and curated data set that incorporates algorithmic developments in image detrending and processing, photometric calibration, and object classification. Y3 GOLD comprises nearly 5000 deg 2 of grizY imaging in the south Galactic cap, including nearly 390 million objects, with depth reaching a signal-to-noise ratio ~10 for extended objects up to iAB ~ 23.0, and top-of-the-atmosphere photometric uniformity <3 mmag. Compared to DR1, photometric residuals with respect to Gaia are reduced by 50%, and per-object chromatic corrections are introduced. Y3 GOLD augments DES DR1 with simultaneous fits to multi-epoch photometry for more robust galactic color measurements and corresponding photometric redshift estimates. Y3 GOLD features improved morphological star–galaxy classification with efficiency >98% and purity >99% for galaxies with 19 < i AB < 22.5. Additionally, it includes per-object quality information, and accompanying maps of the footprint coverage, masked regions, imaging depth, survey conditions, and astrophysical foregrounds that are used to select the cosmologic analysis samples.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2021

Detection of Anomalies in Gamma Background Radiation Data with K-Means and Self-Organizing Map Clustering Algorithms (Consortium on Nuclear Security Technologies (CONNECT) Q1 Report)

Environmental screening of gamma radiation consists of detecting weak nuisance and anomaly signal in the presence of strong and highly varying background. In a typical scenario, a mobile detector-spectrometer continuously measures gamma radiation spectra in short, e.g., one-second, signal acquisition intervals. The measurement data is a 2D matrix, where one dimension is gamma ray energy, and the other dimension is the number of measurements or total time. In principle, gamma radiation sources can be detected and identified from the measured data by their unique spectral lines. Detecting sources from data measured in a search scenario is difficult due to the highly varying background because of naturally occurring radioactive material (NORM), and low signal-to-noise ratio (S/N) of spectral signal measured during one-second acquisition intervals. The objective of this work is to explore unsupervised machine learning (ML) algorithms for detection and identification of weak nuisances and anomalies events in the presence of highly fluctuating background. The challenge is that spectral lines of isotopes are difficult to observe in one-second measurements. Averaging over the entire measurement campaign data set reveals spectral lines of most common background isotopes. Spectral lines of orphan sources, which might appear only in a few measurements during the campaign, will be washed out if averaging is performed over the entire measurement data set. The approach we have explored consists of extracting one-second measurements containing weak spectral features through data clustering. Averaging one-second spectra in a cluster should reveal the presence of anomaly sources. We created two ML models using K-means clustering and Neural Network Self-organizing Map (SOM). Performance of these ML models was benchmarked using search data. One data set contained 137 Cs source, and another dataset contained 131 I source.

61 RADIATION PROTECTION AND DOSIMETRY↗
DOE OSTI2024

Offshore application of landslide susceptibility mapping using gradient-boosted decision trees: a Gulf of Mexico case study

Abstract Among natural hazards occurring offshore, submarine landslides pose a significant risk to offshore infrastructure installations attached to the seafloor. With the offshore being important for current and future energy production, there is a need to anticipate where future landslide events are likely to occur to support planning and development projects. Using the northern Gulf of Mexico (GoM) as a case study, this paper performs Landslide Susceptibility Mapping (LSM) using a gradient-boosted decision tree (GBDT) model to characterize the spatial patterns of submarine landslide probability over the United States Exclusive Economic Zone (EEZ) where water depths are greater than 120 m. With known spatial extents of historic submarine landslides and a Geographic Information System (GIS) database of known topographical, geomorphological, geological, and geochemical factors, the resulting model was capable of accurately forecasting potential locations of sediment instability. Results of a permutation modelling approach indicated that LSM accuracy is sensitive to the number of unique training locations with model accuracy becoming more stable as the number of training regions was increased. The influence that each input feature had on predicting landslide susceptibility was evaluated using the SHapely Additive exPlanations (SHAP) feature attribution method. Areas of high and very high susceptibility were associated with steep terrain including salt basins and escarpments. This case study serves as an initial assessment of the machine learning (ML) capabilities for producing accurate submarine landslide susceptibility maps given the current state of available natural hazard-related datasets and conveys both successes and limitations.

Dyer, Alec S. (ORCID:0000000219813904)↗
DOE OSTI2026

The Emission and Suppression of Line Features in Luminous Transients

Featureless optical and ultraviolet (UV) spectra are a puzzling signature to emerge from recent observations of luminous fast blue optical transients (LFBOTs) and some tidal disruption events (TDEs). We describe the landscape of source and gas properties that are expected to form H, He I , and He II emission lines, and map spectral types to the parameter space of luminosity and system radius. Using one-dimensional radiative transfer calculations, we show that high source luminosities (L > 10 44 erg s −1 ) and compact ejecta radii (r < 10 14 cm) produce featureless spectra, due to the high temperature and ionization state of the emitting medium. Intermediate luminosities and moderately compact systems can generate He II -dominated spectra, while lower luminosities and more extended atmospheres result in conspicuous H and He I emission. Large expansion velocities (v ≥ 0.1c) can further broaden lines such that they blend into the continuum. Featureless UV spectra may require even more extreme ionization environments or velocities in order to suppress the many intrinsically strong metal lines at those wavelengths. Applying this framework to understand the absence of features observed in LFBOTs and featureless TDEs, we find that, under the optically thick, quasi-thermal conditions considered here, nonhomologous, compact outflows are likely necessary for featurelessness to persist in optical and UV spectra.

Aspegren, Olivia [University of California, Berkel↗
DOE OSTI2021

Target space entanglement in Matrix Models

We study target space entanglement in gauged multi-matrix models as models of entanglement between groups of D-branes separated by a planar entangling surface, paying close attention to the implementation of gauge invariance. We open with a review of target space entanglement between identical particles, which shares some important features (specifically a gauged permutation symmetry) with our main problem. For our matrix models, we implement a gauge fixing well-adapted to the entangling surface. In this gauge, we map the matrix model problem to that of entanglement of a U(1) gauge theory on a complete or all-to-all lattice. Matrix elements corresponding to open strings stretching across the entangling surface in the target space lead to interesting contributions to the entanglement entropy.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2022

Multiview Incomplete Knowledge Graph Integration with application to cross-institutional EHR data harmonization

Objective: The growing availability of electronic health records (EHR) data opens opportunities for integrative analysis of multi-institutional EHR to produce generalizable knowledge. A key barrier to such integrative analyses is the lack of semantic interoperability across different institutions due to coding differences. We propose a Multiview Incomplete Knowledge Graph Integration (MIKGI) algorithm to integrate information from multiple sources with partially overlapping EHR concept codes to enable translations between healthcare systems. Methods: The MIKGI algorithm combines knowledge graph information from (i) embeddings trained from the co-occurrence patterns of medical codes within each EHR system and (ii) semantic embeddings of the textual strings of all medical codes obtained from the Self-Aligning Pretrained BERT (SAPBERT) algorithm. Due to the heterogeneity in the coding across healthcare systems, each EHR source provides partial coverage of the available codes. MIKGI synthesizes the incomplete knowledge graphs derived from these multi-source embeddings by minimizing a spherical loss function that combines the pairwise directional similarities of embeddings computed from all available sources. MIKGI outputs harmonized semantic embedding vectors for all EHR codes, which improves the quality of the embeddings and enables direct assessment of both similarity and relatedness between any pair of codes from multiple healthcare systems. Results: With EHR co-occurrence data from Veteran Affairs (VA) healthcare and Mass General Brigham (MGB), MIKGI algorithm produces high quality embeddings for a variety of downstream tasks including detecting known similar or related entity pairs and mapping VA local codes to the relevant EHR codes used at MGB. Based on the cosine similarity of the MIKGI trained embeddings, the AUC was 0.918 for detecting similar entity pairs and 0.809 for detecting related pairs. For cross-institutional medical code mapping, the top 1 and top 5 accuracy were 91.0% and 97.5% when mapping medication codes at VA to RxNorm medication codes at MGB; 59.1% and 75.8% when mapping VA local laboratory codes to LOINC hierarchy. When trained with 500 labels, the lab code mapping attained top 1 and 5 accuracy at 77.7% and 87.9%. MIKGI also attained best performance in selecting VA local lab codes for desired laboratory tests and COVID-19 related features for COVID EHR studies. Compared to existing methods, MIKGI attained the most robust performance with accuracy the highest or near the highest across all tasks. Conclusions: The proposed MIKGI algorithm can effectively integrate incomplete summary data from biomedical text and EHR data to generate harmonized embeddings for EHR codes for knowledge graph modeling and cross-institutional translation of EHR codes.

Zhou, Doudou↗