Search

While deep learning models have made incredible progress across a variety of machine learning tasks, they remain vulnerable to adversarial examples crafted to fool otherwise trustworthy models. Previous work has proposed examining the internal activation of Imagenet models to detect adversarial examples. Our work expands the scale and scope of previous research by simultaneously probing every activation within an Imagenet model using a novel probe block. This probe block model is trained against multiple adversarial algorithms to create a more robust detector. Parameterization of the probe block and adversarial classification networks that utilize probe block output are examined in an ablation experiment with probes of Resnet-50, Inception-v3 and Xception. Considered adversarial classification networks include examples built with Mobilenet-v2 which is shown to be better than a VGG alternative for detecting adversarial artifacts. Results are compared to logistic regression feature squeezing results, which we suggest is an improvement to feature squeezing.

Primary hemophagocytic lymphohistiocytosis is a rare syndrome characterized by immune dysregulation and hyperinflammation. It typically manifests in infancy and is associated with high mortality.We investigated the efficacy and safety of emapalumab (a human anti-interferon-γ antibody), administered with dexamethasone, in an open-label, single-group, phase 2-3 study involving patients who had received conventional therapy before enrollment (previously treated patients) and previously untreated patients who were 18 years of age or younger and had primary hemophagocytic lymphohistiocytosis. The patients could enter a long-term follow-up study until 1 year after allogeneic hematopoietic stem-cell transplantation or until 1 year after the last dose of emapalumab, if transplantation was not performed. The planned 8-week treatment period could be shortened or extended if needed according to the timing of transplantation. The primary efficacy end point was the overall response, which was assessed in the previously treated patients according to objective clinical and laboratory criteria.At the cutoff date of July 20, 2017, a total of 34 patients (27 previously treated patients and 7 previously untreated patients) had received emapalumab; 26 patients completed the study. A total of 63% of the previously treated patients and 65% of the patients who received an emapalumab infusion had a response; these percentages were significantly higher than the prespecified null hypothesis of 40% (P = 0.02 and P = 0.005, respectively). In the previously treated group, 70% of the patients were able to proceed to transplantation, as were 65% of the patients who received emapalumab. At the last observation, 74% of the previously treated patients and 71% of the patients who received emapalumab were alive. Emapalumab was not associated with any organ toxicity. Severe infections developed in 10 patients during emapalumab treatment. Emapalumab was discontinued in 1 patient because of disseminated histoplasmosis.Emapalumab was an efficacious targeted therapy for patients with primary hemophagocytic lymphohistiocytosis. (Funded by NovImmune and the European Commission; NI-0501-04 and NI-0501-05 ClinicalTrials.gov numbers, NCT01818492 and NCT02069899.).

ObjectiveEpi Evident is a web based application built to empower public health analysts by providing a platform that improves monitoring, comparing, and forecasting case counts and period prevalence of notifiable diseases for any scale jurisdiction at regional, country, or global-level. This proof of concept application development addresses improving visualization, access, situational awareness, and prediction of disease behavior.IntroductionThe Epi Evident application was designed for clear and comprehensive visualization for monitoring, comparing, and forecasting notifiable diseases simultaneously across chosen countries. Epi Evident addresses the taxing analytical evaluation of how diseases behave differently across countries. This application provides a user-friendly platform with easily interpretable analytics which allows analysts to conduct biosurveillance with minimal user tasks. Developed at the Pacific Northwest National Laboratory (PNNL), Epi Evident utilizes time-series disease case count data from the Biosurveillance Ecosystem (BSVE) application Epi Archive (1). This diverse data source is filtered through the flexible Epi Evident workflow for forecast model building designed to integrate any entering combination of country and disease. The application aims to quickly inform analysts of anomalies in disease & location specific behavior and aid in evidence based decision making to help control or prevent disease outbreaks.MethodsA workflow was constructed to define the best disease forecast model for each location based on an adjustable method approach. The differences in disease behavior across countries was achieved through a React/Python application with a user-friendly output for monitoring and comparing different combinations.The forecast model building workflow consisted of three major steps to determine the best fit model for a given disease-country pair: data type, model type, and model comparison & selection. Testing various disease-country combinations allowed for direct evaluation of the workflow efficiency, flexibility, and criteria for determining the best fit model. Data type was characterized as either seasonal, cyclic, or sporadic. Depending on data type, a specific time series forecasting model was applied. In general, seasonal or cyclic data required either an Auto-Regression Integrated Moving Average (ARIMA) model or a Seasonal Auto-Regression Integrated Moving Average (SARIMA) model while sporadic datasets employed a Poisson model. Several model candidates for a single country and disease combination were then compared to determine which was the best fit model. ARIMA and SARIMA model selection criteria included their respective order significance, residual diagnostics, and lowest possible combination of Akaike Information Criterion and Root Mean Square Error (RMSE) values. Poisson model selection criteria involved Poisson or negative binomial distribution and event probability, lag dependency of immediate past events or seasonality, and lowest possible RMSE. To enhance the user’s monitoring and comparisons across multiple countries and diseases, each forecasted case counts supplied a corresponding period prevalence. This period prevalence was calculated by dividing the case counts by the population in the selected country and timeframe. Population records were obtained through the public World Health Organization database (2).ResultsA variety of visualization tools on Epi Evident allows convenient interpretation on behaviors of diseases spanning multiple countries simultaneously (Figure 1). Countries, diseases, and timeframe are selected and displayed within a matrix alongside with their corresponding forecasts for case counts and period prevalence. By providing this full representation, users can easily interpret and anticipate disease behavior while monitoring, comparing, and forecasting case counts and period prevalence across multiple countries. For future work, the Epi Evident workflow can be scaled to accommodate any disease-country combination with automated model selection to allow easier and more efficient biosurveillance.ConclusionsEpi Evident empowers analysts to visualize, monitor, compare, and forecast disease case counts and period prevalence across countries. Epi Evident exemplifies how filtering diverse data through a flexible workflow can be scalable to output distinctive models for any given country and disease combination. Thus, providing accurate forecasting and enhanced situational awareness throughout the globe. Implementing this application’s methodology helps enhance and expand biosurveillance efficacy for multiple diseases across multiple countries simultaneously.References1. Generous Nicholas, Fairchild Geoffrey, Khalsa Hari, Tasseff Byron, Arnold James. Epi Archive: An automated data collection of notifiable disease data. Online Journal of Public Health Informatics. 2017. 9(1):e372. http://apps.who.int/gho/data/view.main.POP2040?lang=en Accessed: 6/20/2017

Objective: The Wearable Sensor Application developed by Pacific Northwest National Laboratory (PNNL) provides an early warning system for stressors to individual and group health using physiologic and environmental indicators. The application integrates health monitoring parameters from wearable sensors, e.g., temperature and heart rate, with relevant environmental parameters, e.g., weather and landscape data, and calculates the corresponding physiological strain index. The information is presented to the analyst in a group and individual view with real-time alerting of abnormal health parameters. This application is the first of its kind being developed for integration into the Defense Threat Reduction Agency's Biosurveillance Ecosystem (BSVE). Introduction: Wearable devices are a low cost, minimally invasive way to monitor health. Sensor data provides real-time physiological indictors of an individual’s health status without the requirement of health care professionals or facilities. Information gleamed from wearable sensors can be used to better understand physiological stressors and prodromal symptoms. In addition, this data can be used to monitor individuals that are in high risk of health-related problems. However, raw data from wearable sensors can be overwhelming to process and laborious to monitor for an individual and, even more so, for a group of individuals. Often specific combination of ranges of sensor readings are indicative of changes to health status and need to be evaluated together or used to calculate specific signal parameters. In addition, the environment surrounding the individual needs to be considered when interpreting the data. To address these issues, PNNL has developed an application that collects, analyzes, and integrates wearable sensor data with geographic landscape and weather information to provide a real-time early alert and situational awareness tool for monitoring the health of groups and individuals. Methods: The prototype application described here was a product of PNNL’s BSVE Application Development Competition. The final product that will be deployed in the BSVE is currently under development by PNNL and will vary slightly in the exact design and architecture described. Data. Wearable sensor data was collected from the Rim2Rim (R2R) Watch Study of individuals hiking the Grand Canyon in Arizona [1]. Weather information was obtained from nearby weather stations and mapping features were derived from Google Maps. Calculations. A physiological Strain Index (PSI) was calculated using core temperature estimates derived through a Kalman Filter approach and heart rate [2,3]. Application. The prototype backend application development was based in Python with a MongoDB. The front-end development was built using a scalable architecture and modular approach with components in React and D3. Results: A prototype application was developed this past summer through the PNNL BSVE App Competition (Fig 1). The application was aimed at visualizing wearable sensor data from the Grand Canyon R2R hike dataset. Simulated real-time analysis was used to calculate health status of individuals hiking based on measured physiological parameters and to alert to individuals with signs of physiologic health stress. Visualization tools were incorporated to enable sensor data for individuals and the group to be viewed simultaneously along with pertinent weather, geographic, and elevation data. Many features described in the prototype application will be incorporated into the final BSVE application. The key changes will be 1) the ability to select given time periods for viewing historical data as well as the real-time data collection, 2) environmental data and map view will come from BSVE internal data sources, and 3) the alerts will provide more information and have their own page for reviewing. Conclusions: The Wearable Sensor Application developed by PNNL for integration into the BSVE provides an early warning system for individual and group health using physiologic and environmental parameters. The application highlights health status from wearable sensors and relevant environmental parameters while monitoring a calculated physiological strain index. With this tool, an analyst can easily monitor the health of individuals and groups with the aid of real-time alerting tool for early detection of abnormal health parameters.

Sensalert is a wearable health tracking application that enables monitoring of a target population's health status in real-time. The application integrates health monitoring parameters from wearable sensors, e.g., temperature and heart rate, with relevant environmental parameters, e.g., weather and topography, and calculates the corresponding physiological strain index for each point in time. The objective of this application is to provide early warning, support decision making, and increase situational awareness by displaying comprehensive information and providing alerts at the right time and place. This application is the first of its kind being developed for integration into the Defense Threat Reduction Agencies Bio Surveillance Ecosystem (BSVE). We present the user requirements and design principles for developing this type of application. We hope this work helps and inspires future research in the area of wearable technologies for surveillance.

Public health surveillance systems gain significant benefits from integrating existing early incident detection systems, supported by closed data sources, with public available data. However, identifying potential alerting incidents relies on finding accurate, reliable sources and presenting the high volume of data in a way that increases analysts work efficiency; a challenge for any system that leverages publicly available data. In this paper, we present the design concept and the applied design science research methodology of ChemVeillance, a chemical analyst surveillance system. Our work portrays a system design and approach that translates theoretical design methodology into practice creating a powerful surveillance system built for specific use cases. Researchers, designers, developers, and related professionals in the health surveillance community can build upon the principles and methodology described here to enhance and broaden current surveillance systems leading to improved situational awareness based on a robust integrated early warning system.

Objective Pacific Northwest National Laboratory hosted an intern-based web application development contest in the summer of 2016 centered around developing novel chemical surveillance applications to aid in health situational awareness. Making up the three teams were three graduate students (n=9) from various US schools majoring in non- public health domains, such as computer sicence and user design. The interns suc- cessfully developed three applications that demonstrated a value-add to chemical surveillance—ChemAnalyzer (text analytics), RetroSpect (retrospective analysis of chemical events), and ToxicBusters (geo-based trend analytics). These applications will be the basis for the first chemical surveillance application to be incorporated into the DTRA Biosurveillance Ecosystem (BSVE). Introduction Pacific Northwest National Laboratory (PNNL), on behalf the Defense Threat Reduction Agency (DTRA; project number CB10190), hosts an annual intern- based web app development contest. Previous competitions have focused on mobile biosurveillance applications. The 2016 competition pivoted away from biosurveillance to focus on addressing challenges within the field of chemical surveillance and increasing public health chemical situational awareness. The result of the app will be integrated within the DTRA BSVE. Methods PNNL hosted nine graduate interns for a 10-week period in the summer of 2016 as participants in a summer web application development contest. Students were drawn from such fields as software engineering and user experience and design and placed into three teams of three students. The challenge presented to the interns was to design and develop a fully-functional web application that would address a critical need within the chemical surveillance community. The interns developed their own ideas (vetted by PNNL and DTRA), discovered and inte- grated their own data sources, and produced their own visualizations and an- alytics, independent of any assistence outside of that provided in an advisory capacity. The competition end with a judging event with a panel of subject matter experts and cash awards were distributed to the teams. Results Each team produced a unique application. Although there was mild overlap between some of the ideas, the applications were developed independently and each reflected the unique contributions of the teams. ChemAnalyzer is a text-analytics platform designed to facilitate more data- driven decision, given a corpus of text data about a chemical event. Their plat- form provided the ability to automatically identify and highlight key words in documents related to chemical events. The keywords are drawn from an on- tology installed with the system, as well as any user-identified keywords. The ChemAnalyzer team finished in third place. The RetroSpect team developed a visual analytic tool for performing retrospec- tive analysis and monitoring of chemical events. Their app provided the ability to search and analyze past events, as well as visualization of state and county information for the recorded chemical events. The RetroSpect team finished in second place. The Toxicbusters team—the winners of the competition—created a geo-based situational awareness tool for tracking chemical events. Their app featured an updateable map overlay, search functionality for finding specific or related events, incident and city/state/national-level statistics and trends, as well as news and social media integration based on keywords related to chemical surveillance. Conclusions Each of the apps developed by the teams provides value to an analyst tasked with monitoring chemical events. The apps integrated unique data sources to provides a full picture of a chemical event, and its effects upon the surrounding population. This integrated analytics provides a valuable benefit over existing workflows, where analysts must monitor news, social, and other information sources manually for real-time information. The apps developed by these interns are designed to enable identification and analysis of the incident as quickly as possible, allowing for more timely assessments of the incident and its impacts. The web app development contest provided a unique opportunity for students to learn about the emerging needs in chemical surveillance as it relates to health sit- uational awareness. Students were drawn from a variety of fields and were tasked with developing novel web apps addressing some of the most pressing challenges in the field of chemical surveillance. The ideas generated by the students will help form the basis for future chemical surveillance application development to be integrated with the DTRA BSVE.

Fine cell granularity in modern climate models can produce terabytes of data in each snapshot, causing significant I/O overhead. To address this issue, a method of reducing the I/O latency of high-resolution climate models by identifying and selectively outputting regions of interest is presented. Working with a global cloud-resolving model and running with up to 10,240 processors on a Cray XE6, this method provides significant I/O bandwidth reduction depending on the frequency of writes and the size of the region of interest. The implementation challenges of determining global parameters in a strictly core-localized model and properly formatting output files that only contain subsections of the global grid are addressed, as well as the overall bandwidth impact and benefits of the method. The gains in I/O throughput provided by this method allow dual output rates for high-resolution climate models: a low-frequency global snapshot as well as a high-frequency regional snapshot when events of particular interest occur.

Numerical ensemble forecasting is a powerful tool that drives many risk analysis efforts and decision making tasks. These ensembles are composed of individual simulations that each uniquely model a possible outcome for a common event of interest: e.g., the direction and force of a hurricane, or the path of travel and mortality rate of a pandemic. This paper presents a new visual strategy to help quantify and characterize a numerical ensemble's predictive uncertainty: i.e., the ability for ensemble constituents to accurately and consistently predict an event of interest based on ground truth observations. Our strategy employs a Bayesian framework to first construct a statistical aggregate from the ensemble. We extend the information obtained from the aggregate with a visualization strategy that characterizes predictive uncertainty at two levels: at a global level, which assesses the ensemble as a whole, as well as a local level, which examines each of the ensemble's constituents. Through this approach, modelers are able to better assess the predictive strengths and weaknesses of the ensemble as a whole, as well as individual models. We apply our method to two datasets to demonstrate its broad applicability.

In "white powder incidents" or other suspicious and risky situations relating to deadly diseases or chemicals (e.g., Ebola investigation), those who handle the potentially hazardous materials are the heroes who spearhead the first responder's operations. Although well trained, these heroes need to manage complex problems and make life-or-death decisions while handling the unknown and dangerous. We are motivated to explore how Google Glass can facilitate those heroes' missions. To this end, we conducted contextual inquiry on six biohazard-handling, Personal Protective Equipment (PPE)-wearing professionals. With an inductive thematic analysis, we summarized the heroes' workflow and four groups of "Heroes' Problems". A unique "A3 Model" (Awareness, Analysis, Action) was generated to encapsulate our qualitative findings and proposed Glass features. The findings serve as the groundwork for our future development.

Like many local governments, Los Angeles County has been re-shaped profoundly in recent years by a series of external forces such as economic downturns and technological advances. At the same time, the value of strategic human resources management has been increasingly recognized by practitioners and management researchers. Partly to help respond to the external forces, the Los Angeles County Training Academy instituted a “Building a Strategic Human Resources Partnership” program designed to build the competencies needed for HR managers to serve as true strategic partners. This article (a) reports on the decision to build a strategic human resources competency model — rather than to modify or adopt an existing competency model — as a basis for the design of the training program, (b) details the methodologies used to build a reliable, customized list of strategic HR competencies and a categorization structure, (c) describes the outcomes and (d) reports on implications for other local government agencies.

“Color Theory for Automotive Coatings” is a state-of-the-art learning approach to the world of automotive color. This learning tool provides important information to design engineers, color formulators, chemists, paint shop managers, color technicians, and anyone who wants to be more discerning about the science and communication of color. The content synthesizes both historic and current relevant color knowledge bases. The interactive educational design incorporates concepts of adult learning. The fundamentals of color and its application to the automotive industry are presented at the student's own pace. The learner has control of the learning process. © 2003 Wiley Periodicals, Inc. Col Res Appl, 28, 327–334, 2003; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/col.10176

Study objectives: To determine whether the adenine (A)-guanine (G) substitution polymorphism at position - 308 on the tumor necrosis factor-a gene confers susceptibility to COPD or to the development of a more severe form of disease. Design: A cross-sectional study was undertaken to compare the frequency of the A allele in a group of 106 patients with COPD with that in a control population (n 5 99). Patients were followed up prospectively for a period of 2 years. Participants and setting: Participants included 106 COPD patients recruited from a respiratory outpatient clinic and 99 control subjects recruited from patients admitted for cardiac catheterization. Measurements and results: DNA was extracted from venous blood, and each subject was genotyped for the polymorphism by polymerase chain reaction amplification and restriction digestion using Nco1. There was no increased frequency of the A allele in patients compared to control subjects. AA homozygous patients had less reversible airflow obstruction (p < 0.05) and a significantly greater mortality (both all-cause and respiratory deaths) on follow-up (p < 0.001), despite a shorter cigarette smoking history. Conclusions: This study suggests that homozygosity for this A allele predisposes to more severe airflow obstruction and a worse prognosis in COPD. (CHEST 2000; 118:971‐975)

The purpose of the lessons learned document for the BEOWulf Biosurveillance Mobile App Development Intern Competition is to capture the project’s lessons learned in a formal document for use by other project managers on similar future projects. This document may be used as part of new project planning for similar projects in order to determine what problems occurred and how those problems were handled and may be avoided in the future. Additionally, this document details what went well with the project and why, so that other project managers may capitalize on these actions. Project managers may also use this document to determine who the project team members were in order to solicit feedback for planning their projects in the future. This document will be formally communicated with the organization and will become a part of the organizational assets and archives.

OBJECTIVE: To report a case of pulmonary hypersensitivity associated with furazolidone use and review the literature on this topic. CASE SUMMARY: A 43-year-old white female presented with fever and dyspnea. She had recently completed a course of furazolidone 125 mg 4 times daily for 10 days for enteritis. Investigations revealed bibasilar interstitial infiltrates on chest X-ray, hypoxia, and 21% eosinophilia. Her fever, hypoxia, and dyspnea rapidly abated following discontinuation of furazolidone and administration of corticoteroids. DISCUSSION: Furazolidone is a bactericidal agent used to treat infectious enteropathies. It is chemically similar to nitrofurantoin, which is well known to cause pulmonary hypersensitivity reactions. Application of the Naranjo probability scale suggests that a furazolidone adverse reaction in this patient was probable. A MEDLINE search from 1966 to October 2004 revealed 2 previously reported cases suggestive of furazolidone pulmonary hypersensitivity. All published reports closely resemble each other and descriptions of nitrofurantoin-associated pulmonary hypersensitivity reactions. CONCLUSIONS: Furazolidone may induce pulmonary hypersensitivity reactions; clinicians should be aware of this potentially serious adverse effect.

In FY2013, the PIE International Safeguards team demonstrated our development progress to U.S. Department of Energy (DOE) staff from the Office of Nonproliferation and International Security (NA-24, our client) and the Office of Defense Nuclear Nonproliferation Research and Development (NA-22). Following the demonstration, the team was asked by our client to complete additional development prior to a planned demonstration at the International Atomic Energy Agency (IAEA), scheduled tentatively for January or spring of 2014. The team discussed four potential areas for development (in priority order), and will develop them as time and funding permit prior to an IAEA demonstration. The four capability areas are: 1. Addition of equipment manuals to PIE-accessible files 2. Optical character recognition (OCR) of photographed text 3. Barcode reader with information look-up from a database 4. Add Facilities to Data Model 5. Geospatial capabilities with information integration Each area will be described below in a use case.

When querying a huge image database containing millions of images, the result of the query may still contain many thousands of images that need to be presented to the user. We consider the problem of arranging such a large set of images into a visually coherent layout, one that places similar images next to each other. Image similarity is determined using a bag-of-features model, and the layout is constructed from a hierarchical clustering of the image set by mapping an in-order traversal of the hierarchy tree into a space-filling curve. This layout method provides strong locality guarantees so we are able to quantitatively evaluate performance using standard image retrieval benchmarks. Performance of the bag-of-features method is best when the vocabulary is learned on the image set being clustered. Because learning a large, discriminative vocabulary is a computationally demanding task, we present a novel method for efficiently adapting a generic visual vocabulary to a particular dataset. We evaluate our clustering and vocabulary adaptation methods on a variety of image datasets and show that adapting a generic vocabulary to a particular set of images improves performance on both hierarchical clustering and image retrieval tasks.

Author(s): Harivandi, M Ali; Baird, James; Hartin, Janet; Henry, Michael; Shaw, David | Abstract: With long, hot, dry summers common to most areas of the state, California lawns generally need watering for much of the year. Learn how to use water more efficiently as water demands and costs go up and drought conditions continue.

Author(s): Harivandi, M Ali; Gibeault, Victor A; Henry, Michael J; Wu, Lin; Geisel, Pamela M; Unruh, Carolyn L | Abstract: Turfgrass species and varieties rated from best to worst for 14 different criteria involving climate, disease resistance, and intended use of the lawn.

To the Editors: Schiff and colleagues' report (1) on preventable factors in inpatient theophylline toxicity documents the many circumstances that may contribute to the development of theophylline t...