Abstract

Abstract
Several image classification problems are handled using a classical statistical pattern recognition methodology: image segmentation, visual feature extraction, classification. The accuracy of the solution is typically measured by comparing automatic results with manual classification ones, where the distinction between these three steps is not clear at all. In this paper we will focus on one of these steps by addressing the following question: does the visual relevance exploited by segmentation algorithms reflect the semantic relevance of the manual annotation performed by the user? For this purpose we chose a gastroenterology scenario where clinicians classified a set of images into three different types (cancer, pre-cancer, normal), and manually segmented the area they believe was responsible for this classification. Afterwards, we have quantified the performance of two popular segmentation algorithms (mean shift, normalized cuts) on how well they produced one image patch that approximates manual annotation. Results showed that, for this case study, this resemblance is quite close for a large percentage of the images when using normalized cuts.

Abstract
This paper analyzes experimentally the limitations observed in lightwave systems using distributed Raman amplifiers operating under large pump power input conditions. The Brillouin effect is observed as the pump power becomes equal to watt centered in a specific pump wavelength. The presence of Brillouin peaks degrades the system performance. The Raman amplifier is set with singlemode and dispersion compensating fibers and evaluated in a high capacity WDM link at transmission rates up to 40 Gb/s per channel. The dispersion compensation was accomplished by adjusting the length of the dispersion compensating fiber. The system and amplifier characterization were developed in terms of the on-off gain, ripple, optical signal-to-noise ratio, eye diagram and extinction ratio.

Abstract
Despite all the differences on markets, cultural environment, and organizational objectives, some changes are modifying the way business occurs, turning competitive advantages difficult to accomplish. In a global marketplace scenario, changes are constant and with variable amplitude. Organizations need to be alert to all movements to counter-answer and re-align themselves with the new market conditions. Thus, it is wise to develop a flexible strategy, where it is possible to monitor the market conditions and the overall performance of the organization. We propose a project portfolio management and evaluation framework which gathers important concepts and tools from different areas, like Organizational Strategy, Organizational Learning, Decision Support Systems (DSS), Project Portfolio Management and Organizational Maturity, fundamental for acquiring and developing competitive advantages, and, therefore, for increasing the organization's overall performance.

Abstract
This paper describes the present state of evolution of the "Brain Imaging Network Grid" (BING) project in Portugal. This network is composed of 5 nodes connected through a Gbps dark fiber bitways and is served by two clusters and (-)60 TBytes of storage. A distributed information system to support operations and manage the BING community is already in place and further developments to evolve the infrastructure to provide computational neurosciences services are under way and described in the present paper. Moreover, we undercover the architecture that we are implementing that includes the intention to add BING resources to the EGEE node IEETA/UA has been exploring since 2005 under the EGEE Biomed VO. Other aspects of the BING infrastructure evolution aimed to constitute a Grid virtual community in brain imaging are presented and discussed.

Abstract
A smart laminated composite with a simple geometry was produced with embedded fiber Bragg grating sensors. The smart structure is composed of two Bragg gratings located in regions of different thicknesses of the laminated composite. Although one of the Bragg gratings is embedded between two layers, the other is embedded in jour layers. When the strain is applied to the smart composite, different response is obtained. Because of this characteristic it is possible to discriminate strain and temperature using a traditional matrix method. To reduce the inherent error, an artificial neural network approach is proposed that will improve the strain and temperature measurement discrimination when using this new configuration. These instrumented carbon fiber laminates can be used for monitoring of reinforcement and protection of structures. (C) 2008 Wiley Periodicals, Inc. Microwave Opt Technol Lett 51: 235-239, 2009; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/mop.23990