Abstract
Introduction: Pharyngocutaneous fistula after larynx and hypopharynx cancer surgery can cause serveral damages. This study's aim was to derive a clinical decision rule to predict pharyngocutaneous fistula development after pharyngolaryngeal cancer surgery. Methods: A retrospective cohort study was conducted, including all patients performing total laryngectomy/pharyngolaryngectomy (n = 171). Association between pertinent variables and pharyngocutaneous fistula development was assessed and a predictive model proposed. Results: American Society of Anesthesiologists scale, chemoradiotherapy, and tracheotomy before surgery were associated with fistula in the univariate analysis. In the multivariate analysis, only American Society of Anesthesiologists maintained statistical significance. Using logistic regression, a predictive model including the following was derived: American Society of Anesthesiologists, alcohol, N-classification, and diabetes mellitus. The model's score area under the curve was 0.76 (95% CI 0.64-0.87). The high-risk group presented specificity of 93%, positive likelihood ratio of 7.10, and positive predictive value of 76%. Including the medium-low, medium-high, and high-risk groups, a sensitivity of 92%, negative likelihood ratio of 0.25, and negative predictive value of 89% were observed. Conclusion: A clinical decision rule was created to identify patients with high risk of pharyngocutaneous fistula development. Prognostic accuracy measures were substantial. Nevertheless, it is essential to conduct larger prospective studies for validation and refinement.

Abstract
Humans sense the surrounding environment through five senses: sight, hearing, taste, smell and touch. This is valid not only for real life but also when one is exposed to virtual environments. It already exists systems that exploit the sensorial stimulation of more than one sense, however there are no available low-cost solutions that allow creative professionals to create and visualize multisensory contents. On this paper it is presented a solution for the creation and delivery of multisensory contents that allow users to create their own multisensory experiments. An evaluation of the developed solution that shows the potential of these new contents is also described.

Abstract
In this paper, two techniques for interrogation and multiplexing of fiber loop mirror (FLM) intensity sensors based on optical time domain reflectometer (OTDR) are proposed. These configurations enable series and parallel FLM sensor interrogation. A fiber taper characterized as a displacement sensor was used as the intensity sensor. The OTDR parameters were optimized in order to obtain the best results. The optimized parameters were 100-ns pulse width and 10-dB input signal attenuation which permitted to attain similar to 18 dB dynamic range in the operating wavelength of 1550 nm. The results show a linear behavior for both configurations with similar slope, -15.3 dB/mm, in the normalized displacement range of 0.2 to 0.7 mm. It was also achieved a displacement resolution of 0.027 and 0.093 mm, for the series and parallel configurations, respectively. Sensors multiplexing are demonstrated for both configurations and the systems do not present crosstalk. Based on the experimental results, the best configuration is the parallel one. The proposed approach is a viable alternative for multiplexing and interrogation of remote fiber sensors.

Abstract
This work presents the use of a genetic algorithm to design a Mandani Fuzzy Controller with two inputs and one output, written in Matlab (R) environment, applied to a two axis positioning system using a robot. The robot has 6 degrees of freedom and is controlled with the objective of capturing an object on a workspace using a fuzzy controller. A genetic algorithm is used in order to determine the main characteristics of the membership functions of the fuzzy controller. The complete system employed to simulate the two axes positioning system uses the Transfer Function of two axes of the robot and the Fuzzy controller. In this work was implemented and simulated an operating scenario, being the results and the performance of the controller presented regarding the controller energy effort and the evolution of the (x,y) trajectories over time.

Abstract

Abstract
The aggregation of network traffic has been shown to enhance the performance of wireless sensor networks. By reducing the number of packets that are transmitted, energy consumption, collisions and congestion are minimised. However, current data aggregation schemes restrict developers to a specific network structure or cannot handle multi-hop data aggregation. In this paper, we propose Hitch Hiker, a remote component binding model that provides support for multi-hop data aggregation. Hitch Hiker uses component meta-data to discover remote component bindings and to construct a multi-hop overlay network within the free payload space of existing traffic ows. This overlay network provides end-To-end routing of low-priority traffic while using only a small fraction of the energy of standard communication. We have developed a prototype implementation of Hitch Hiker for the LooCI component model. Our evaluation shows that Hitch Hiker consumes minimal resources and that using Hitch Hiker to deliver low-priority traffic reduces energy consumption by up to 15%.