Using neural networks to select wavelet features for breast cancerdiagnosis

This study focuses on improving microcalcification classification by establishing an efficient computer-aided diagnosis system that extracts Daubechies-4 and biorthogonal wavelet features. These wavelets were chosen because they have been used in military target recognition and fingerprint recognition research with images characterized by low contrast, similar to mammography. Feature selection techniques are employed to further increase classification performance. The artificial neural network feature selection techniques are complemented by a conventional decision boundary-based feature selection method. The results using the wavelet features are compared to more conventional measures of image texture, angular second moment, and Karhunen Loeve coefficients. The use of alternative signal processing to compare wavelet and neural techniques allows for a measure of the problem difficulty. It is concluded that advances and contributions have been made with the introduction of two novel feature extraction methods for breast cancer diagnosis, wavelets and eigenmasses. Additionally, feature selection techniques are demonstrated, compared, and validated, transforming adequate discrimination power into promising classification results     

Long-term information collection with energy harvesting wireless sensors: a multi-armed bandit based approach

This paper reports on the development of a multi-agent approach to long-term information collection in networks of energy harvesting wireless sensors. In particular, we focus on developing energy management and data routing policies that adapt their behaviour according to the energy that is harvested, in order to maximise the amount of information collected given the available energy budget. In so doing, we introduce a new energy management technique, based on multi-armed bandit learning, that allows each agent to adaptively allocate its energy budget across the tasks of data sampling, receiving and transmitting. By using this approach, each agent can learn the optimal energy budget settings that give it efficient information collection in the long run. Then, we propose two novel decentralised multi-hop algorithms for data routing. The first proveably maximises the information throughput in the network, but can sometimes involve high communication cost. The second algorithm provides near-optimal performance, but with reduced computational and communication costs. Finally, we demonstrate that, by using our approaches for energy management and routing, we can achieve a 120% improvement in long-term information collection against state-of-the-art benchmarks.     

HEAT TRANSFER BETWEEN ROUGH SURFACES IN CONTACT OVER A HIGHLY ELLIPTICAL CONTOUR AREA: COMPARISON OF EXPERIMENTAL AND NUMERICAL RESULTS

This paper describes experimental and numerical studies of heat transfer between surfaces in contact over a highly elliptical area in a gas or in vacuum. Measurements of the steady-state thermal resistance of a high aspect ratio contact formed between cylindrical steel surfaces with widely different radii of curvature (76 cm vs 0.95 cm) are reported for two levels of surface roughness (0.05 μm and 1.0 μm rms) and compared to 3-D numerical results obtained with the multigrid method. Theoretical results obtained by evaluating the contact conductance acting over each surface element within the contour area with a method developed previously for rough but nominally flat surfaces are shown to be in excellent agreement with the rough surface experimental data.     

Stability of nonlinear 2D systems described by the continuous-time Roesser model

This paper considers systems with two-dimensional dynamics (2D systems) described by the continuous-time nonlinear state-space Roesser model. The sufficient conditions of exponential stability in terms of vector Lyapunov functions are established. These conditions are then applied to analysis of the absolute stability of a certain class of systems comprising a linear continuous-time plant in the form of the Roesser model with a nonlinear characteristic in the feedback loop, which satisfies quadratic constraints. The absolute stability conditions are reduced to computable expressions in the form of linear matrix inequalities. The obtained results are extended to the class of continuous-time systems governed by the Roesser model with Markovian switching. The problems of absolute stability and stabilization via state- and output-feedback are solved for linear systems of the above class. The solution procedures for these problems are in the form of algorithms based on linear matrix inequalities.     

The composition of volatile oils derived from oleoresins

Essential oils and oleoresins derived from spices and herbs, by steam distillation in the former case or by solvent extraction in the latter case, are increasingly more important as flavoring constituents in heat-processed foods. Steam-distilled oils, when used for such purposes, tend to steam-distill during the heat processing due to moisture that is present. Oleoresins that contain the essential oil and natural fixatives tend to depress this volatilization and are preferred as flavoring materials in such cases. Subjective flavor panel evaluations are reported, showing the significance between oleoresins and essential oils as well as the corresponding spice or herb. Instrumental evaluation, primarily gas chromatography, will compare essential oils produced by direct steam distillation vs. those isolated form oleoresins. The difference in composition which can be related to the improved flavor effect from oleoresins is discussed. One of 13 papers presented in the symposium “Flavor Research in Fats and Fat Bearing foods,” AOCS Meeting, Atlantic City, October 1971.     

The Effects of Experience and Strategy on Visual Attention Allocation in an Automated Multiple-Task Environment

Operators and users interacting with computer environments often have to deal with multiple tasks at once, responding to each in series. Diagnostic automation, that is, automation that alerts users when and where to look, has been suggested to support the unique challenges of multiple task environments: activating tasks, switching between tasks, and tasks interfering with each other. Automation is not always reliable, however. Because of the common interaction with novel systems and the importance of training, the Simultaneous Task Environment Platform program—a multiple-task environment—was developed to understand the effects of experience on interaction with these automation-supported systems, as well as what strategies were developed. It was found that participants became more efficient with experience only when they interacted with higher reliability automation. Furthermore, the strategies participants developed focused on the differences between tasks and patterns across those tasks. Automated systems training should be sure to employ these findings.     

Agent Technologies for Sensor Networks

Wireless sensor networks are increasingly seen as a solution to the problem of performing continuous wide-area monitoring in many environmental, security, and military scenarios. The distributed nature of such networks and the autonomous behavior expected of them present many novel challenges. In this article, the authors argue that a new synthesis of electronic engineering and agent technology is required to address these challenges, and they describe three examples where this synthesis has succeeded. In more detail, they describe how these novel approaches address the need for communication and computationally efficient decentralized algorithms to coordinate the behavior of physically distributed sensors, how they enable the real-world deployment of sensor agent platforms in the field, and finally, how they facilitate the development of intelligent agents that can autonomously acquire data from these networks and perform information processing tasks such as fusion, inference, and prediction.     

Effects of Environmental Oxygen Content and Dissolved Oxygen on the Surface Tension and Viscosity of Liquid Nickel

The NASA Marshall Space Flight Center’s electrostatic levitation (ESL) laboratory has recently added an oxygen partial pressure controller. This system allows the oxygen partial pressure within the vacuum chamber to be measured and controlled in the range from approximately \(10^{-28}\,{\mathrm {to}}\,10^{-9}\) bar, while in a vacuum atmosphere. The oxygen control system installed in the ESL laboratory’s main chamber consists of an oxygen sensor, oxygen pump, and a control unit. The sensor is a potentiometric device that determines the difference in oxygen activity in two gas compartments (inside the chamber and the air outside of the chamber) separated by an electrolyte. The pump utilizes coulometric titration to either add or remove oxygen. The system is controlled by a desktop control unit, which can also be accessed via a computer. The controller performs temperature control for the sensor and pump, has a PID-based current loop and a control algorithm. Oxygen partial pressure has been shown to play a significant role in the surface tension of liquid metals. Oxide films or dissolved oxygen may lead to significant changes in surface tension. The effects on surface tension and viscosity by oxygen partial pressure in the surrounding environment and the melt dissolved oxygen content will be evaluated, and the results will be presented. The surface tension and viscosity will be measured at several different oxygen partial pressures while the sample is undercooled. Surface tension and viscosity will be measured using the oscillating droplet method.     

High throughput research and evaporation rate modeling for solvent screening for ethylcellulose barrier membranes in pharmaceutical applications

Context: Ethylcellulose is commonly dissolved in a solvent or formed into an aqueous dispersion and sprayed onto various dosage forms to form a barrier membrane to provide controlled release in pharmaceutical formulations. Due to the variety of solvents utilized in the pharmaceutical industry and the importance solvent can play on film formation and film strength it is critical to understand how solvent can influence these parameters.

Objective: To systematically study a variety of solvent blends and how these solvent blends influence ethylcellulose film formation, physical and mechanical film properties and solution properties such as clarity and viscosity.

Materials and methods: Using high throughput capabilities and evaporation rate modeling, thirty-one different solvent blends composed of ethanol, isopropanol, acetone, methanol, and/or water were formulated, analyzed for viscosity and clarity, and narrowed down to four solvent blends. Brookfield viscosity, film casting, mechanical film testing and water permeation were also completed.

Results and discussion: High throughput analysis identified isopropanol/water, ethanol, ethanol/water and methanol/acetone/water as solvent blends with unique clarity and viscosity values. Evaporation rate modeling further rank ordered these candidates from excellent to poor interaction with ethylcellulose. Isopropanol/water was identified as the most suitable solvent blend for ethylcellulose due to azeotrope formation during evaporation, which resulted in a solvent-rich phase allowing the ethylcellulose polymer chains to remain maximally extended during film formation. Consequently, the highest clarity and most ductile films were formed.

Conclusion: Employing high throughput capabilities paired with evaporation rate modeling allowed strong predictions between solvent interaction with ethylcellulose and mechanical film properties.     

Metal microparticle – Polymer composites as printable,bio/ecoresorbable conductive inks

Biologically and environmentally resorbable electronic devices support application possibilities that cannot be addressed with conventional technologies. This paper presents highly conductive, water-soluble composites that can be printed to form contacts, interconnects, antennas, and other important features that are essential to nearly all systems of this type. An optimized material formulation involves in situ polymerization to yield a polyanhydride containing a dispersion of molybdenum microparticles at appropriate concentrations. Comparisons of essential physical and electrical properties of these materials to those of composites formed with other polymers and other metal microparticles reveal the relevant considerations. Various functional demonstrations of screen-printed test structures and devices illustrate the suitability of these conductive inks for use in water-soluble electronic devices. A key advantage of the material introduced here compared to alternatives is its ability to maintain conductance over significant periods of time while immersed in relevant aqueous solutions. Studies involving live animal models establish the biocompatibility.