Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “biological material”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 289 records · Page 16
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Chorioallantoic membrane vascularization. A meta-analysis

The CAM is a widely used experimental assay to study angiogenesis, wound healing, tumor growth and metastatic process. In this study, we have analyzed and compared the existent literature data concerning the growth of the CAM. Moreover, we have analyzed the data concerning the development of the vascular system and the expression of the most important pro-angiogenic and anti-angiogenic factors. The availability of these data and their comparative evaluation allow to better analyze the experimental data concerning the testing of different pro-angiogenic and anti-angiogenic molecules, as well as biomaterials in the CAM assay. Moreover, the dynamic of the angiogenic response to different tumor cell lines and or tumor bioptic specimens, may be also better evaluated and estimated.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2021

Decellularized ECM derived from normal bone involved in the viability and chemo-sensitivity in multiple myeloma cells

Multiple myeloma (MM) is an incurable plasma cell malignancy. The progression of MM is closely related to the bone microenvironment. Bone matrix proteins are remodeled and manipulated to govern cancer growth during the process of MM. However the role of normal bone extracellular matrix in MM is still unclear. In this study the decellularized extracellular matrix derived from normal SD rats’ skulls (N-dECM) was prepared by decellularization technology. The CCK 8 assay and the dead-live cell kit assay were used to determine the viability of MM cells and the sensitivity to bortezomib. The Realtime PCR and Western blot assay were used to assay the mRNA and protein related to MM. Under the treatment of N-dECM, we found that the viability of MM cells was inhibited and the sensitivity of MM cells to bortezomib was increased. Additionally, the expression levels of APRIL and TACI, which participated in the progression of MM, were significantly decreased in MM cells. It suggested that N-dECM might inhibit the development of MM via APRIL-TACI axis, and our study may provide a novel and potential biomaterial for MM therapy.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2021

Injectable platelet rich fibrin facilitates hair follicle regeneration by promoting human dermal papilla cell proliferation, migration, and trichogenic inductivity

Hair follicle regeneration has been successful in mice but failed in human being for years. Dermal papilla cells, a specialized mesenchymal stem cell derived from dermal papilla within hair follicles, is considered the key cells for hair follicle regeneration function as both regeneration initiator and regulator. Injectable platelet rich fibrin (i-PRF), a novel biomaterial rich in a variety of growth factors and three-dimensional scaffolds, has shown promising effects on tissue regeneration. In this study, we aimed to evaluate the application of i-PRF in human hair follicle regeneration by examining the biological effects of i-PRF on human dermal papilla cells (hDPCs). Biomaterial compatibility, cell viability, proliferation, migration, alkaline phosphatase activity and trichogenic inductivity were assessed after exposing hDPCs to different concentrations of i-PRF extracts. In addition, we investigated the ultrastructure of i-PRF with all cell components filtered. The results revealed that i-PRF possessing excellent biocompatibility and could significantly promote hDPCs proliferation, migration, and trichogenic inductivity. Furthermore, the concentration of i-PRF is able to remarkably influence hDPCs behavior in a dose-dependent pattern. Different concentrations exhibited differential effects on hDPCs behavior. In general, lower concentration promotes cell proliferation better than higher concentration, while higher concentration promotes cell function better reversely. Best concentration for hDPCs in vitro expending is 1% concentration. 20% concentration is optimal for hair follicle regeneration. In summary, our findings concluded that i-PRF facilitates hair follicle regeneration by promoting human dermal papilla cell proliferation, migration, and trichogenic inductivity.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2020

An Open-Cell Environmental Transmission Electron Microscopy Technique for In Situ Characterization of Samples in Aqueous Liquid Solutions

The desire to image specimens in liquids has led to the development of open-cell and closed-cell techniques in transmission electron microscopy (TEM). The closed-cell approach is currently more common in TEM and has yielded new insights into a number of biological and materials processes in liquid environments. The open-cell approach, which requires an environmental TEM (ETEM), is technically challenging but may be advantageous in certain circumstances due to fewer restrictions on specimen and detector geometry. Here, we demonstrate a novel approach to open-cell liquid TEM, in which we use salt particles to facilitate the in situ formation of droplets of aqueous solution that envelope specimen particles coloaded with the salt. This is achieved by controlling sample temperature between 1 and 10°C and introducing water vapor to the ETEM chamber above the critical pressure for the formation of liquid water on the salt particles. Our use of in situ hydration enables specimens to be loaded into a microscope in a dry state using standard 3 mm TEM grids, allowing specimens to be prepared using trivial sample preparation techniques. Furthermore, our future aim will be to combine this technique with an in situ light source to study photocorrosion in aqueous environments.

47 OTHER INSTRUMENTATION↗
DOE OSTI2020

Liquid–Liquid Phase Separation in Supermicrometer and Submicrometer Aerosol Particles

The interactions of aerosol particles with light and clouds are among the most uncertain aspects of anthropogenic climate forcings. The effects of aerosol particles on climate depend on their optical properties, heterogeneous chemistry, water uptake behavior, and ice nucleation activity. These properties in turn depend on aerosol physics and chemistry including composition, size, shape, internal structure (morphology), and phase state. The greatest numbers of particles are found at small, submicrometer sizes, and the properties of aerosol particles can differ on the nanoscale compared with measurements of bulk materials. As a result, our focus has been on characterizing the phase transitions of aerosol particles in both supermicrometer and submicrometer particles. The phase transition of particular interest for us has been liquid–liquid phase separation (LLPS), which occurs when components of a solution phase separate due to a difference in solubilities. For example, organic compounds can have limited solubility in salt solutions especially as the water content decreases, increasing the concentration of the salt solution, and causing phase separation between organic-rich and inorganic-rich phases. To characterize the systems of interest, we primarily use optical microscopy for supermicrometer particles and cryogenic-transmission microscopy for submicrometer particles. This Account details our main results to date for the phase transitions of supermicrometer particles and the morphology of submicrometer aerosol. Herein, we have found that the relative humidity (RH) at which LLPS occurs (separation RH; SRH) is highly sensitive to the composition of the particles. For supermicrometer particles, SRH decreases as the pH is lowered to atmospherically relevant values. SRH also decreases when non-phase-separating organic compounds are added to the particles. For submicrometer particles, a size dependence of morphology is observed in systems that undergo LLPS in supermicrometer particles. In the limit of slow drying rates, particles <30 nm are homogeneous and larger particles are phase-separated. This size dependence of aerosol morphology arises because small particles cannot overcome the activation barrier needed to form a new phase when phase separation occurs by a nucleation and growth mechanism. Additionally, the inhibition of LLPS in small particles is observed for mixtures of ammonium sulfate with single organic compounds as well as complex organics like α-pinene secondary organic matter. The morphology of particles affects activation diameters for the formation of cloud condensation nuclei. These results more generally have implications for aerosol properties that affect the climate system. In addition, LLPS is also widely studied in materials and biological chemistry, and our results could potentially translate to implications for these fields.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2026

Mechanistic Insights into Dinitrogen Reduction to Ammonia in Light-Controlled Nanocrystal:Nitrogenase Complexes

Developing systems that can efficiently capture photon energy and convert this energy into fuels and chemicals requires understanding how to assemble molecular components with diverse functions into complete systems possessing selectivity and efficiency in directing charge carriers to catalytic reactions. There are many challenges to achieving this goal. One promising approach is the development of hybrid systems that combine semiconductor nanocrystals (NCs) for light capture and enzymes as efficient catalysts. Such biohybrid systems capitalize on the tunable electronic and optical properties of NCs while leveraging the unmatched specificity and efficiency of enzymes in catalyzing chemical reactions, thereby offering opportunities to surpass the limitations of each component alone. Here, we focus on recent progress in developing a biohybrid system that combines CdS NCs for photon capture with the enzyme nitrogenase to accomplish light-driven dinitrogen (N 2 ) reduction to ammonia (NH 3 ). Integrating light-harvesting materials with biological catalysts requires a deep understanding of NC properties, protein stability, and electron transfer (ET), making it an inherently multidisciplinary problem. The reduction of N 2 to NH 3 is a challenging reaction, with a high demand in both agriculture and industrial chemical production. This reaction is intrinsically energy intensive, due to the need to activate the N≡N triple bond. The current standard industrial approach to N 2 reduction, the Haber−Bosch reaction, obtains the necessary energy input from fossil fuels, whereas biological systems capable of N 2 reduction utilize the hydrolysis of ATP as their energy source. Replacing these costly, energy-intensive inputs with renewable light energy represents a critical step toward sustainable NH 3 production. Recent progress has demonstrated that semiconductor CdS NCs can be coupled to the catalytic component of nitrogenase, the MoFe protein, to form a biohybrid CdS NC:MoFe protein complex, enabling light-driven N 2 reduction rather than energy input from fossil fuels or ATP. This illustrates how inorganic NCs can functionally replace the natural Fe protein partner, yielding a biohybrid catalyst that enables controlled electron delivery and provides not only light-driven NH 3 production but also new approaches for probing enzyme catalytic function. The CdS NC:MoFe protein biohybrid system enables light-initiated electron delivery at ambient temperature, as well as temperatures below freezing, allowing for stabilization and spectroscopic characterization of key reaction intermediates. These findings highlight how photochemical biohybrids can serve as both functional catalysts and mechanistic probes. Beyond studies of the nitrogenase mechanism, studies of the CdS NC:MoFe system reveal how variables such as NC size, electrostatic binding interactions, and sacrificial electron donors (SEDs) govern complex stability, charge transfer efficiency, and catalytic performance. In addition, studies of nitrogenase and the high activation barrier for N 2 reduction are enabling investigations of new and interesting questions regarding the properties and limitations of NC biocatalysis. In this Account, we describe the key features of CdS NC:MoFe protein biohybrids and the parameters for optimal light-driven N 2 reduction, and how controlling ET with light illuminates the path to new insights into the nitrogenase mechanism.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

Deep Learning-Enabled MS/MS Spectrum Prediction Facilitates Automated Identification Of Novel Psychoactive Substances

The market for illicit drugs has been reshaped by the emergence of more than 1100 new psychoactive substances (NPS) over the past decade, posing a major challenge to the forensic and toxicological laboratories tasked with detecting and identifying them. Tandem mass spectrometry (MS/MS) is the primary method used to screen for NPS within seized materials or biological samples. The most contemporary workflows necessitate labor-intensive and expensive MS/MS reference standards, which may not be available for recently emerged NPS on the illicit market. Here, we present NPS-MS, a deep learning method capable of accurately predicting the MS/MS spectra of known and hypothesized NPS from their chemical structures alone. NPS-MS is trained by transfer learning from a generic MS/MS prediction model on a large data set of MS/MS spectra. We show that this approach enables a more accurate identification of NPS from experimentally acquired MS/MS spectra than any existing method. We demonstrate the application of NPS-MS to identify a novel derivative of phencyclidine (PCP) within an unknown powder seized in Denmark without the use of any reference standards. We anticipate that NPS-MS will allow forensic laboratories to identify more rapidly both known and newly emerging NPS. NPS-MS is available as a web server at https://nps-ms.ca/, which provides MS/MS spectra prediction capabilities for given NPS compounds. Additionally, it offers MS/MS spectra identification against a vast database comprising approximately 8.7 million predicted NPS compounds from DarkNPS and 24.5 million predicted ESI-QToF-MS/MS spectra for these compounds.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

Machine Learning Prediction of the Critical Cooling Rate for Metallic Glasses from Expanded Datasets and Elemental Features

In this study, we use a random forest (RF) model to predict the critical cooling rate (R C ) for glass formation of various alloys from features of their constituent elements. The RF model was trained on a database that integrates multiple sources of direct and indirect R C data for metallic glasses to expand the directly measured R C database of less than 100 values to a training set of over 2000 values. The model error on 5-fold cross-validation (CV) is 0.66 orders of magnitude in K/s. The error on leave-out-one-group CV on alloy system groups is 0.59 log units in K/s when the target alloy constituents appear more than 500 times in training data. Using this model, we make predictions for the set of compositions with melt-spun glasses in the database and for the full set of quaternary alloys that have constituents which appear more than 500 times in training data. These predictions identify a number of potential new bulk metallic glass systems for future study, but the model is most useful for the identification of alloy systems likely to contain good glass formers rather than detailed discovery of bulk glass composition regions within known glassy systems.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Molecular Design of Functional Polymers for Silica Scale Inhibition

Silica polymerization, which involves the condensation reaction of silicic acid, is a fundamental process with wide-ranging implications in biological systems, material synthesis, and scale formation. The formation of a silica-based scale poses significant technological challenges to energy-efficient operations in various industrial processes, including heat exchangers and water treatment membranes. Despite the common strategy of applying functional polymers for inhibiting silica polymerization, the underlying mechanisms of inhibition remain elusive. In this study, we synthesized a series of nitrogen-containing polymers as silica inhibitors and elucidated the role of their molecular structures in stabilizing silicic acids. Polymers with both charged amine and uncharged amide groups in their backbones exhibit superior inhibition performance, retaining up to 430 ppm of reactive silica intact for 8 h under neutral pH conditions. In contrast, monomers of these amine/amide-containing polymers as well as polymers containing only amine or amide functionalities present insignificant inhibition. Molecular dynamics simulations reveal strong binding between the deprotonated silicic acid and a polymer when the amine groups in the polymer are protonated. Notably, an extended chain conformation of the polymer is crucial to prevent proximity between the interacting monomeric silica species, thereby facilitating effective silica inhibition. Furthermore, the hydrophobic nature of alkyl segments in polymer chains disrupts the hydration shell around the polymer, resulting in enhanced binding with ionized silicic acid precursors compared to monomers. Finally, our findings provide novel mechanistic insights into the stabilization of silicic acids with functional polymers, highlighting the molecular design principles of effective inhibitors for silica polymerization.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2021

Deep Learning Coordinate-Free Quantum Chemistry

Computing quantum chemical properties of small molecules and polymers can provide insights valuable to physicists, chemists and biologists when designing new materials, catalysts, biological probes and drugs. Deep learning can compute quantum chemical properties accurately in a fraction of the time required by commonly used methods such as density functional theory (DFT). However, many of these deep learning architectures require energy minimized molecular geometries as input, which is also computationally expensive, and decreasing the reproducibility and throughput of these methods. In this study, we demonstrate that accurate quantum chemical computations can be performed without optimized geometries by operating in the coordinate-free domain using deep learning on graph encodings. Furthermore, we also find that the choice of graph-encoding architecture substantially affects the performance of these methods. The Wave architecture outperforms graph convolution architectures, particularly on complex molecules. Furthermore, the structures of these graph encoding architectures provide an opportunity to probe an important, outstanding question in quantum mechanics: What types of quantum chemical properties can be represented by local-variable models? We find that Wave, a local-variable model, is more accurately calculates quantum chemical properties. Graph convolutional architectures require global variables, and are not as effective as as Wave. We anticipate that coordinate-free, deep-learning models of quantum chemistry will become valuable tools in chemistry and biology, enabling researchers to rapidly screen chemical databases or identify new molecules using automated, de-novo design algorithms.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

Understanding the Structure and Dynamics of Complex Biomembrane Interactions by Neutron Scattering Techniques

The membrane is one of the key structural materials of biology at the cellular level. Composed predominantly of a bilayer of lipids with embedded and bound proteins, it defines the boundaries of the cell and many organelles essential to life and therefore is involved in almost all biological processes. Membrane-specific interactions, such as drug binding to a membrane receptor or the interactions of an antimicrobial compound with the lipid matrix of a pathogen membrane, are of interest across the scientific disciplines. Herein we present a review, aimed at nonexperts, of the major neutron scattering techniques used in membrane studies: small-angle neutron scattering, neutron membrane diffraction, neutron reflectometry, quasielastic neutron scattering, and neutron spin echo. Neutron scattering techniques are well suited to studying biological membranes. The nondestructive nature of cold neutrons means that samples can be measured for long periods without fear of beam damage from ultraviolet, electron, or X-ray radiation, and neutron beams are highly penetrating, thus offering flexibility in samples and sample environments. Most important is the strong difference in neutron scattering lengths between the two most abundant forms of hydrogen, protium and deuterium. Changing the relative amounts of protium/deuterium in a sample allows the production of a series of neutron scattering data sets, enabling the observation of differing components within complex membrane architectures. This approach can be as simple as using the naturally occurring neutron contrast between different biomolecules to study components in a complex by changing the solution H 2 O/D 2 O ratio or as complex as selectively labeling individual components with hydrogen isotopes. This review presents an overview of each experimental technique with the neutron instrument configuration, related sample preparation and sample environment, and data analysis, highlighted by a special emphasis on using prominent neutron contrast to understand structure and dynamics. This review gives researchers a practical introduction to the often enigmatic suite of neutron beamlines, thereby lowering the barrier to taking advantage of these large-facility techniques to achieve new understandings of membranes and their interactions with other molecules.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2024

Molecular Mechanisms of Sorbed Ion Effects during Boehmite Particle Aggregation

Classical theories of particle aggregation, such as Derjaguin–Landau–Verwey–Overbeek (DLVO), do not explain recent observations of ion-specific effects or the complex concentration dependence for aggregation. Thus, here, we probe the molecular mechanisms by which selected alkali nitrate ions (Na + , K + , and NO 3 –) influence aggregation of the mineral boehmite (γ-AlOOH) nanoparticles. Nanoparticle aggregation was analyzed using classical molecular dynamics (CMD) simulations coupled with the metadynamics rare event approach for stoichiometric surface terminations of two boehmite crystal faces. Calculated free energy landscapes reveal how electrolyte ions alter aggregation on different crystal faces relative to pure water. Consistent with experimental observations, we find that adding an electrolyte significantly reduces the energy barrier for particle aggregation (~3–4×). However, in this work, we show this is due to the ions disrupting interstitial water networks, and that aggregation between stoichiometric (010) basal–basal surfaces is more favorable than between (001) edge–edge surfaces (~5–6×) due to the higher interfacial water densities on edge surfaces. The interfacial distances in the interlayer between aggregated particles with electrolytes (~5–10 Å) are larger than those in pure water (a few Ångströms). Together, aggregation/disaggregation in salt solutions is predicted to be more reversible due to these lower energy barriers, but there is uncertainty on the magnitudes of the energies that lead to aggregation at the molecular scale. By analyzing the peak water densities of the first monolayer of interstitial water as a proxy for solvent ordering, we find that the extent of solvent ordering likely determines the structures of aggregated states as well as the energy barriers to move between them. Further, the results suggest a path for developing a molecular-level basis to predict the synergies between ions and crystal faces that facilitate aggregation under given solution conditions. Such fundamental understanding could be applied extensively to the aggregation and precipitation utilization in the biological, pharmaceutical, materials design, environmental remediation, and geological regimes.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Superhydrophobic Array Devices for the Enhanced Formation of 3D Cancer Models

During the metastatic cascade, cancer cells travel through the bloodstream as circulating tumor cells (CTCs) to a secondary site. Clustered CTCs have greater shear stress and treatment resistance, yet their biology remains poorly understood. We therefore engineered a tunable superhydrophobic array device (SHArD). The SHArD-C was applied to culture a clinically relevant model of CTC clusters. Using our device, we cultured a model of cancer cell aggregates of various sizes with immortalized cancer cell lines. These exhibited higher E-cadherin expression and are significantly more capable of surviving high fluid shear stress-related forces compared to single cells and model clusters grown using the control method, helping to explain why clustering may provide a metastatic advantage. Additionally, the SHArD-S, when compared with the AggreWell 800 method, provides a more consistent spheroid-forming device culturing reproducible sizes of spheroids for multiple cancer cell lines. Overall, we designed, fabricated, and validated an easily tunable engineered device which grows physiologically relevant three-dimensional (3D) cancer models containing tens to thousands of cells.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

A Nitrogen Out-of-Plane (NOOP) Mechanism for Imine-Based Light-Driven Molecular Motors

Light-driven molecular motors have generated considerable interest due to their potential applications in material and biological systems. Recently, Greb and Lehn reported a new class of molecular motors, chiral N-alkyl imines, which undergo unidirectional rotation induced by light and heat. The mechanism of unidirectional motion in molecular motors containing a C═N group has been assumed to consist of photoinduced torsion about the double bond. In this work, we present a computational study of the photoisomerization dynamics of a chiral N-alkyl imine motor. We find that the location and energetics of minimal energy conical intersections (MECIs) alone are insufficient to understand the mechanism of the motor. Furthermore, a key part of the mechanism consists of out-of-plane distortions of the N atom (followed by isomerization about the double bond). Dynamic effects and out-of-plane distortions are critical to understand the observed (rather low) quantum yield for photoisomerization. Finally, our results provide hints as to how the photoisomerization quantum yield might be increased, improving the efficiency of this class of molecular motors.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

Revealing core-valence interactions in solution with femtosecond X-ray pump X-ray probe spectroscopy

Abstract Femtosecond pump-probe spectroscopy using ultrafast optical and infrared pulses has become an essential tool to discover and understand complex electronic and structural dynamics in solvated molecular, biological, and material systems. Here we report the experimental realization of an ultrafast two-color X-ray pump X-ray probe transient absorption experiment performed in solution. A 10 fs X-ray pump pulse creates a localized excitation by removing a 1 s electron from an Fe atom in solvated ferro- and ferricyanide complexes. Following the ensuing Auger–Meitner cascade, the second X-ray pulse probes the Fe 1 s → 3 p transitions in resultant novel core-excited electronic states. Careful comparison of the experimental spectra with theory, extracts +2 eV shifts in transition energies per valence hole, providing insight into correlated interactions of valence 3 d with 3 p and deeper-lying electrons. Such information is essential for accurate modeling and predictive synthesis of transition metal complexes relevant for applications ranging from catalysis to information storage technology. This study demonstrates the experimental realization of the scientific opportunities possible with the continued development of multicolor multi-pulse X-ray spectroscopy to study electronic correlations in complex condensed phase systems.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2020

Layer-by-Layer Freezing of Nanoconfined Water

Abstract Nanoconfined water plays a pivotal role in a vast number of fields ranging from biological and materials sciences to catalysis, nanofluidics and geochemistry. Here, we report the freezing and melting behavior of water (D 2 O) nanoconfined in architected silica-based matrices including Vycor glass and mesoporous silica SBA-15 and SBA-16 with pore diameters ranging between 4–15 nm, which are investigated using differential scanning calorimetry and 2 H nuclear magnetic resonance spectroscopy. The results provide compelling evidence that the extreme dynamical heterogeneity of water molecules is preserved over distances as small as a few angstroms. Solidification progresses in a layer-by-layer fashion with a coexistence of liquid-like and solid-like dynamical fraction at all temperatures during the transition process. The previously reported fragile-to-strong dynamic transition in nanoconfined water is argued to be a direct consequence of the layer-by-layer solidification.

42 ENGINEERING↗
DOE OSTI2021

Limited-angle computed tomography with deep image and physics priors

Computed tomography is a well-established x-ray imaging technique to reconstruct the three-dimensional structure of objects. It has been used extensively in a variety of fields, from diagnostic imaging to materials and biological sciences. One major challenge in some applications, such as in electron or x-ray tomography systems, is that the projections cannot be gathered over all the angles due to the sample holder setup or shape of the sample. This results in an ill-posed problem called the limited angle reconstruction problem. Typical image reconstruction in this setup leads to distortion and artifacts, thereby hindering a quantitative evaluation of the results. To address this challenge, we use a generative model to effectively constrain the solution of a physics-based approach. Our approach is self-training that can iteratively learn the nonlinear mapping from partial projections to the scanned object. Because our approach combines the data likelihood and image prior terms into a single deep network, it is computationally tractable and improves performance through an end-to-end training. We also complement our approach with total-variation regularization to handle high-frequency noise in reconstructions and implement a solver based on alternating direction method of multipliers. We present numerical results for various degrees of missing angle range and noise levels, which demonstrate the effectiveness of the proposed approach.

47 OTHER INSTRUMENTATION↗
DOE OSTI2022

Scalable and accurate multi-GPU-based image reconstruction of large-scale ptychography data

Abstract While the advances in synchrotron light sources, together with the development of focusing optics and detectors, allow nanoscale ptychographic imaging of materials and biological specimens, the corresponding experiments can yield terabyte-scale volumes of data that can impose a heavy burden on the computing platform. Although graphics processing units (GPUs) provide high performance for such large-scale ptychography datasets, a single GPU is typically insufficient for analysis and reconstruction. Several works have considered leveraging multiple GPUs to accelerate the ptychographic reconstruction. However, most of these works utilize only the Message Passing Interface to handle the communications between GPUs. This approach poses inefficiency for a hardware configuration that has multiple GPUs in a single node, especially while reconstructing a single large projection, since it provides no optimizations to handle the heterogeneous GPU interconnections containing both low-speed (e.g., PCIe) and high-speed links (e.g., NVLink). In this paper, we provide an optimized intranode multi-GPU implementation that can efficiently solve large-scale ptychographic reconstruction problems. We focus on the maximum likelihood reconstruction problem using a conjugate gradient (CG) method for the solution and propose a novel hybrid parallelization model to address the performance bottlenecks in the CG solver. Accordingly, we have developed a tool, called PtyGer ( Pty chographic G PU(multipl e )-based r econstruction), implementing our hybrid parallelization model design. A comprehensive evaluation verifies that PtyGer can fully preserve the original algorithm’s accuracy while achieving outstanding intranode GPU scalability.

97 MATHEMATICS AND COMPUTING↗