Catalogue Search | MBRL
Are you sure you want to remove the book from the shelf?
{{itemTitle}}
137
result(s) for
"Day, Alexandra"
Sort by:
Carl at the dog show
Carl the rottweiler explores the dog show in which his brother Gamble competes.
Book
Image processing pipeline for AI-driven nanoparticle megalibrary characterization
Recent innovations have made it possible to produce megalibraries, millions of structurally and compositionally distinct nanoparticles on a chip. These megalibraries yield vast volumes of data that are impossible to analyze manually, necessitating the development of automated tools. In previous work, we created a binary classification machine learning model to select quality nanoparticle images for downstream analysis. In this work, we show that adding a custom image processing step before training can produce significantly higher-performing models in a fraction of the time and make them more robust to different image noise levels and microscope acquisition settings. The image processing pipeline proposed here effectively cleans raw nanoparticle images, enhances key features, and allows us to use much lower resolution images and simpler neural network model architectures. These features result in higher performance and significant cost savings. Experiments demonstrate superior performance relative to baseline, including an 18.2% improvement in recall and a 13.1% increase in accuracy. Given the high cost of downstream analysis, it is critical to minimize false positives, and our best-performing model reaches a precision of 95.9% and a weighted F-score of 95.1% on an unseen test set. Additionally, model training time is reduced from hours to less than a minute. We also show that, using this custom image processing pipeline, model performance is significantly improved at lower pixel resolutions compared to downsizing alone. We expect that adopting this pipeline for AI-driven automated nanoparticle characterization will allow researchers to rapidly and accurately analyze much greater volumes of data, thereby accelerating materials discovery.
Journal Article
Automated image segmentation for accelerated nanoparticle characterization
Recent developments in materials science have made it possible to synthesize millions of individual nanoparticles on a chip. However, many steps in the characterization process still require extensive human input. To address this challenge, we present an automated image processing pipeline that optimizes high-throughput nanoparticle characterization using intelligent image segmentation and coordinate generation. The proposed method can rapidly analyze each image and return optimized acquisition coordinates suitable for multiple analytical STEM techniques, including 4D-STEM, EELS, and EDS. The pipeline employs computer vision and unsupervised learning to remove the image background, segment the particle into areas of interest, and generate acquisition coordinates. This approach eliminates the need for uniform grid sampling, focusing data collection on regions of interest. We validated our approach using a diverse dataset of over 900 high-resolution grayscale nanoparticle images, achieving a 96.0% success rate based on expert-validated criteria. Using established 4D-STEM acquisition times as a baseline, our method demonstrates a 25.0 to 29.1-fold reduction in total processing time. By automating this crucial preprocessing step and optimizing data acquisition, our pipeline significantly accelerates materials characterization workflows while reducing unnecessary data collection.
Journal Article
On a mushroom day
\"Structured around a walk in the woods, On a Mushroom Day is an artful introduction to the wonderful world of mushrooms.\"-- Provided by publisher.
CHILDBOOK
Graph Neural Network for Object Reconstruction in Liquid Argon Time Projection Chambers
This paper presents a graph neural network (GNN) technique for low-level reconstruction of neutrino interactions in a Liquid Argon Time Projection Chamber (LArTPC). GNNs are still a relatively novel technique, and have shown great promise for similar reconstruction tasks in the LHC. In this paper, a multihead attention message passing network is used to classify the relationship between detector hits by labelling graph edges, determining whether hits were produced by the same underlying particle, and if so, the particle type. The trained model is 84% accurate overall, and performs best on the EM shower and muon track classes. The model’s strengths and weaknesses are discussed, and plans for developing this technique further are summarised.
Journal Article
Building a Bridge Beyond the Wall: Transitioning Care From Jail Health Services to a Community Clinic
The Sinclair Health Clinic (SHC) provides services to underinsured residents living in Winchester, Frederick, and Clarke Counties. The SHC was established in 1986 with the mission of improving access to healthcare services, removing barriers to healthcare, and promoting the overall well-being of the community. The Northwestern Regional Adult Detention Center (NRADC) in Frederick County, Virginia was established in 1845, and is now a 540-bed medium security corrections facility. When individuals complete their sentence, they are reentered into the community and are often seen at the SHC for follow up services. No standardized process for this transition from incarceration to community exists resulting in unmet needs. This quality improvement project provides a summary of the current reentry process for individuals facing release from a regional jail in a rural community in northwestern Virginia. The summary will be utilized by key stakeholders to develop protocols that will facilitate a seamless transition for inmates.
Dissertation
Houses of History and Homepages: Museums of Communism in Eastern and Central Europe and their Online Presence
Museums of communism have been appearing all over Central and Eastern Europe, as has the interest in them. Memorabilia sells well and tourists are interested in what life was like behind the Iron Curtain. If these museums show how they present, represent, and deal with the recent past, museum websites also give insight into the museums’ greater outreach. Communism is portrayed in several ways in different countries, but they generally depict the country as a victim of Soviet rule and influence. This study examines resources that are available for the global audience and observes how engaged the museums are in presenting their information via new media and internet websites. Case studies of museums of communism are presented from the following museums and historical education centers: The Stasi Museum and the DDR Museum in Berlin, Germany; the Institute of National Remembrance in Warsaw, Poland; the Museum of Communism in Prague, Czech Republic; the House of Terror Museum and Memento Park in Budapest, Hungary; and the Museum of Occupations in Tallinn, Estonia. These sites of public memory can be seen as metaphors for the historic transition from communism in these respective states. How museums and online displays of history are conceptualized show, in part, how countries and their historians are dealing with the recent past, how that past is presented to the public, and how they want to portray it to the world.
Dissertation
Graph Neural Network for Object Reconstruction in Liquid Argon Time Projection Chambers
This paper presents a graph neural network (GNN) technique for low-level reconstruction of neutrino interactions in a Liquid Argon Time Projection Chamber (LArTPC). GNNs are still a relatively novel technique, and have shown great promise for similar reconstruction tasks in the LHC. In this paper, a multihead attention message passing network is used to classify the relationship between detector hits by labelling graph edges, determining whether hits were produced by the same underlying particle, and if so, the particle type. The trained model is 84% accurate overall, and performs best on the EM shower and muon track classes. The model's strengths and weaknesses are discussed, and plans for developing this technique further are summarised.
Paper
Planarized Fabrication Process With Two Layers of SIS Josephson Junctions and Integration of SIS and SFS {\\pi}-Junctions
We present our new fabrication Process for Superconductor Electronics (PSE2) that integrates two (2) layers of Josephson junctions in a fully planarized multilayer process on 200-mm wafers. The two junction layers can be, e.g., conventional Superconductor-Insulator-Superconductor (SIS) Nb/Al/AlO_x/Nb junctions with the same or different Josephson critical current densities, J_c. The process also allows integration of high-J_c Superconductor-Ferromagnet-Superconductor (SFS) or SFS'S JJs on the first junction layer with Nb/Al/AlO_x/Nb trilayer junctions on the second junction layer, or vice versa. In the present node, the SFS trilayer, Nb/Ni/Nb is placed below the standard SIS trilayer and separated by one niobium wiring layer. The main purpose of integrating the SFS and SIS junction layers is to provide compact {\\pi}-phase shifters in logic cells of superconductor digital circuits and random access memories, and thereby increase the integration scale and functional density of superconductor electronics. The current node of the two-junction-layer process has six planarized niobium layers, two layers of resistors, and 350-nm minimum feature size. The target Josephson critical current densities for the SIS junctions are 100 {\\mu}A/{\\mu}m^2 and 200 {\\mu}A/{\\mu}m^2. We present the salient features of the new process, fabrication details, and characterization results on two layers of Josephson junctions integrated into one process, both for the conventional and {\\pi}-junctions.
Paper