Image Reconstruction Using Interval Simulated Annealing in Electrical Impedance Tomography

Electrical impedance tomography (EIT) is an imaging technique that attempts to reconstruct the impedance distribution inside an object from the impedance between electrodes placed on the object surface. The EIT reconstruction problem can be approached as a nonlinear nonconvex optimization problem in which one tries to maximize the matching between a simulated impedance problem and the observed data. This nonlinear optimization problem is often ill-posed, and not very suited to methods that evaluate derivatives of the objective function. It may be approached by simulated annealing (SA), but at a large computational cost due to the expensive evaluation process of the objective function, which involves a full simulation of the impedance problem at each iteration. A variation of SA is proposed in which the objective function is evaluated only partially, while ensuring boundaries on the behavior of the modified algorithm.

Electroconvulsive Therapy in Brazil After the "Psychiatric Reform" A Public Health Problem-Example From a University Service

Objectives: The Brazilian public health system does not provide electroconvulsive therapy (ECT), which is limited to a few academic services. National mental health policies are against ECT. Our objectives were to analyze critically the public policies toward ECT and present the current situation using statistics from the Institute of Psychiatry of the University of Sao Paulo (IPq-HCFMUSP) and summary data from the other 13 ECT services identified in the country. Methods: Data regarding ECT treatment at the IPq-HCFMUSP were collected from January 2009 to June 2010 (demographical, number of sessions, and diagnoses). All the data were analyzed using SPSS 19, Epic Info 2000, and Excel. Results: During this period, 331 patients were treated at IPq-HCFMUSP: 221 (67%) were from Sao Paulo city, 50 (15.2%) from Sao Paulo's metropolitan area, 39 (11.8%) from Sao Paulo's countryside, and 20 (6.1%) from other states; 7352 ECT treatments were delivered-63.0% (4629) devoted entirely via the public health system (although not funded by the federal government); the main diagnoses were a mood disorder in 86.4% and schizophrenia in 7.3% of the cases. Conclusions: There is an important lack of public assistance for ECT, affecting mainly the poor and severely ill patients. The university services are overcrowded and cannot handle all the referrals. The authors press for changes in the mental health policies.

Molecular chemistry and the missing mass problem in planetary nebulae

Context. Detections of molecular lines, mainly from H-2 and CO, reveal molecular material in planetary nebulae. Observations of a variety of molecules suggest that the molecular composition in these objects differs from that found in interstellar clouds or in circumstellar envelopes. The success of the models, which are mostly devoted to explain molecular densities in specific planetary nebulae, is still partial however. Aims. The present study aims at identifying the influence of stellar and nebular properties on the molecular composition of planetary nebulae by means of chemical models. A comparison of theoretical results with those derived from the observations may provide clues to the conditions that favor the presence of a particular molecule. Methods. A self-consistent photoionization numerical code was adapted to simulate cold molecular regions beyond the ionized zone. The code was used to obtain a grid of models and the resulting column densities are compared with those inferred from observations. Results. Our models show that the inclusion of an incident flux of X-rays is required to explain the molecular composition derived for planetary nebulae. We also obtain a more accurate relation for the N(CO)/N(H-2) ratio in these objects. Molecular masses obtained by previous works in the literature were then recalculated, showing that these masses can be underestimated by up to three orders of magnitude. We conclude that the problem of the missing mass in planetary nebulae can be solved by a more accurate calculation of the molecular mass.

The impact of the medical speciality in primary health-care problem solving in Belo Horizonte, Brazil: homeopaths versus family doctors: a preliminary quantitative study

Introduction: This research project examined influence of the doctors' speciality on primary health care (PHC) problem solving in Belo Horizonte (BH) Brazil, comparing homeopathic with family health doctors (FH), from the management's and the patients' viewpoint. In BH, both FH and homeopathic doctors work in PHC. The index of resolvability (IR) is used to compare resolution of problems by doctors. Methods: The present research compared IR, using official data from the Secretariat of Health and test requests made by the doctors and 482 structured interviews with patients. A total of 217,963 consultations by 14 homeopaths and 67 FH doctors between 1 July 2006 and 30 June 2007 were analysed. Results: The results show significant differences greater problem resolution by homeopaths compared to FH doctors. Conclusion: In BH, the medical speciality, homeopathy or FH, has an impact on problem solving, both from the managers' and the patients' point of view. Homeopaths request fewer tests and have better IR compared with FH doctors. Specialisation in homeopathy is an independent positive factor in problem solving at PHC level in BH, Brazil. Homeopathy (2012) 101, 44-50.

Combining meta-learning and search techniques to select parameters for support vector machines

Support Vector Machines (SVMs) have achieved very good performance on different learning problems. However, the success of SVMs depends on the adequate choice of the values of a number of parameters (e.g., the kernel and regularization parameters). In the current work, we propose the combination of meta-learning and search algorithms to deal with the problem of SVM parameter selection. In this combination, given a new problem to be solved, meta-learning is employed to recommend SVM parameter values based on parameter configurations that have been successfully adopted in previous similar problems. The parameter values returned by meta-learning are then used as initial search points by a search technique, which will further explore the parameter space. In this proposal, we envisioned that the initial solutions provided by meta-learning are located in good regions of the search space (i.e. they are closer to optimum solutions). Hence, the search algorithm would need to evaluate a lower number of candidate solutions when looking for an adequate solution. In this work, we investigate the combination of meta-learning with two search algorithms: Particle Swarm Optimization and Tabu Search. The implemented hybrid algorithms were used to select the values of two SVM parameters in the regression domain. These combinations were compared with the use of the search algorithms without meta-learning. The experimental results on a set of 40 regression problems showed that, on average, the proposed hybrid methods obtained lower error rates when compared to their components applied in isolation.