Questionnaire- versus voice-based screening for laryngeal disorders

The usefulness of questionnaire and voice data to screen for laryngeal disorders is explored. Answers to 14 questions form a questionnaire data vector. Twenty-three variables computed by the commercial “Dr.Speech” software from a digital voice recording of a sustained phonation of the vowel sound/a/constitute a voice data vector. Categorization of the data into a healthy class and two classes of disorders, namely diffuse and nodular mass lesions of vocal folds is the task pursued in this work. Visualization of data and automated decisions is also an important aspect of this work. To make the categorization, a support vector machine (SVM) is designed based on genetic search. Linear as well as nonlinear canonical correlation analysis (CCA) is employed, to study relations between the questionnaire and voice data sets. The curvilinear component analysis, performing nonlinear mapping into a two-dimensional space, is used for visualizing data and decisions. Data from 240 patients were used in the experimental studies. It was found that the questionnaire data provide more information for the categorization than the voice data. There are 3-4 common directions along which the statistically significant variations of the questionnaire and voice data occur. However, the linear relations between the variations occurring in the two data sets are not strong. On the other hand, very strong linear relations were observed between the nonlinear variates obtained from the questionnaire data and linear ones computed from the voice data. Questionnaire data carry great potential for preventive health care in laryngology.     

Multi-resolution HDD contact detection technique for low SNR applications

The accurate calibration of active flying height in a hard disk drive (HDD) is an essential metrology for proper drive operation. Detecting head-disk contact of a head-disk-interface (HDI) is a critical step in calibrating the active fly height. Proposed is a head-disc contact detection technique that is an improved adaptive discrete wavelet transform (ADWT) methodology that can be integrated into an advanced HDD. The method is designed to process low SNR head/disc proximity probe data during a thermal protrusion pulse and extract HDI head modulation signatures correlating them with the contact event. The significance of the proposed ADWT implementation is in fact that thousands of DWT coefficients can be represented by a simple statistically derived data set of four parameters that can be tracked to declare the contact. This compact approach enables a practical, cost-effective ADWT implementation into the HDD hardware architecture using existing micro-processors and on-board memory.     

Quantitative and qualitative lipid correlation in experimental endogenous hyperlipemia

The reversible endogenous hyperlipemia in dogs, elicited by the detergent Triton which was given intravenously, was used to study the interrelations of serum lipids. In the cholesterol ester fraction an increase occurs in both monounsaturated and in saturated fatty acids, excepting myristic; while a decrease occurs in polyunsaturated fatty acids. The fatty acids of cholesterol esters of normal dogs contain 22% oleic acid, and only 24% when serum lipids are increased to almost double their normal value (TC=400–500 mg/100 ml). However there is a critical level above which a rapid rise in oleic acid occurs and, in severe hyperlipemia (TC=1500 ±430 mg/100 ml), this acid constitutes almost half of the esterified fatty acid component. Since there is no evidence that Triton directly regulates fatty acid synthesis, the lipid fraction-fatty acid interrelationship may be secondary to lipid mobilization from endogenous sources. This concept is supported by the fact that the increased serum fatty acids are only those which can be synthesized by animals. It is suggested, on the basis of a marked increased of endogenously produced fatty acids, that, at critical lipid levels, shortage of polyunsaturated fatty acids from exogenous sources occurs. This might be of sufficient degree to accelerate fatty acid synthesis to meet the need for fatty acids for energy requirements. There may also be need of fatty acid for esterification of chiefly the accumulated free cholesterol split from lipoprotein by Triton. Triton-induced changes in cholesterol ester fatty acids result in patterns which closely resemble those in the adipose tissue of dog and man and in the serum of human endogenous hyperlipemia.     

Using Unlabelled Data to Train a Multilayer Perceptron

This Letter presents an approach to using both labelled and unlabelled data to train a multilayer perceptron. The unlabelled data are iteratively pre-processed by a perceptron being trained to obtain the soft class label estimates. It is demonstrated that substantial gains in classification performance may be achieved from the use of the approach when the labelled data do not adequately represent the entire class distributions. The experimental investigations performed have shown that the approach proposed may be successfully used to train neural networks for learning different classification problems.     

A novel approach to designing an adaptive committee applied to predicting company’s future performance

This article presents an approach to designing an adaptive, data dependent, committee of models applied to prediction of several financial attributes for assessing company’s future performance. Current liabilities/Current assets, Total liabilities/Total assets, Net income/Total assets, and Operating Income/Total liabilities are the attributes used in this paper. A self-organizing map (SOM) used for data mapping and analysis enables building committees, which are specific (committee size and aggregation weights) for each SOM node. The number of basic models aggregated into a committee and the aggregation weights depend on accuracy of basic models and their ability to generalize in the vicinity of the SOM node. A random forest is used a basic model in this study. The developed technique was tested on data concerning companies from ten sectors of the healthcare industry of the United States and compared with results obtained from averaging and weighted averaging committees. The proposed adaptivity of a committee size and aggregation weights led to a statistically significant increase in prediction accuracy if compared to other types of committees.     

A general framework for designing a fuzzy rule-based classifier

This paper presents a general framework for designing a fuzzy rule-based classifier. Structure and parameters of the classifier are evolved through a two-stage genetic search. To reduce the search space, the classifier structure is constrained by a tree created using the evolving SOM tree algorithm. Salient input variables are specific for each fuzzy rule and are found during the genetic search process. It is shown through computer simulations of four real world problems that a large number of rules and input variables can be eliminated from the model without deteriorating the classification accuracy. By contrast, the classification accuracy of unseen data is increased due to the elimination.     

Opto-nanomechanical test instrument in mechanical characterization of DLC coated MEMS devices

Nanomechanical test instruments can precisely measure force and displacement and produce repeatable load-unload curves that can be used in characterizing mechanical behaviour at the nanoscale. In this study, a newly developed translucent nanomechanical test instrument was used in deriving reproducible stiffness values for ultra nano crystalline diamond (UNCD)-made atomic force microscope (AFM) cantilevers that are used for mechanical properties characterization of DLC films. The instrument which is calibrated according to the ISO-14577 standard is integrated with nano/micropositioners and multi-view optical microscopes to facilitate precise manipulation of those AFM cantilevers that were dozens of micrometers in size. Experimental results on UNCD-made AFM cantilever stiffness measurements are provided and compared with theoretical and batch production values.

     

Automatic benthic imagery recognition using a hierarchical two-stage approach

The main objective of this work is to establish an automated classification system of seabed images. A novel two-stage approach to solving the image region classification task is presented. The first stage is based on information characterizing geometry, colour and texture of the region being analysed. Random forests and support vector machines are considered as classifiers in this work. In the second stage, additional information characterizing image regions surrounding the region being analysed is used. The reliability of decisions made in the first stage regarding the surrounding regions is taken into account when constructing a feature vector for the second stage. The proposed technique was tested in an image region recognition task including five benthic classes: red algae, sponge, sand, lithothamnium and kelp. The task was solved with the average accuracy of 90.11% using a data set consisting of 4589 image regions and the tenfold cross-validation to assess the performance. The two-stage approach allowed increasing the classification accuracy for all the five classes, more than 27% for the “difficult” to recognize “kelp” class.     

Governmental policy and prospect in electricity production from renewables in Lithuania

In Lithuania, the generation of electricity is based on the nuclear energy and on the fossil fuels. After the decommissioning of Ignalina nuclear power plant in 2009, the Lithuanian Power Plant and other thermal plants will become the major sources of electricity. Consequently, the Lithuanian power sector must focus on the implementation of renewable energy projects, penetration of new technologies and on consideration of the future opportunities for renewables, and Government policy for promoting this kind of energy. Production of electricity from renewable energy is based on hydro, biomass and wind energy resources in Lithuania. Due to the typical climatic condition in Lithuania the solar photovoltaics and geothermal energy are not used for power sector. Moreover, the further development of hydropower plants is limited by environmental restrictions, therefore priority is given to wind energy development and installation of new biomass power plants. According to the requirements set out in the Directive 2001/77/EC of the European Parliament and of the Council of 27 September 2001 on the promotion of electricity produced from renewable energy sources in the internal electricity market [Official Journal L283, 33–40, 27 October 2001], 7% of gross consumption of electricity will be generated from renewable energy by 2010 in Lithuania. The aim of this paper is to show the estimation of the maximum renewable power penetration in the Lithuanian electricity sector and possible environmental impact.     

Effects of diets rich in saturated fatty acids with or without added cholesterol on plasma lipids and lipoproteins

Semi-synthetic diet I which contained 16% hydrogenated coconut oil and 5% cholesterol, and diet II, identical to I but without cholesterol supplement, were fed to dogs for four months to determine the effects of added cholesterol on lipemia produced by diets high in saturated fatty acids (FA) and lacking essential FA. In addition, diet I was fed to another group of dogs for 12 to 16 months. The initiation of lipemia was very similar in all experimental animals. Plasma from dogs on diets I and II showed significant increases in lipid concentration and changes in FA per cent composition within the first week, as compared to controls, while during the first month there was no difference in lipid concentration or FA distribution in all lipid fractions between I and II. At the 10th and 16th weeks plasma total and free cholesterol and phospholipid were significantly higher in the group on diet I, with the cholesterol supplement, than on diet II with no added cholesterol, but there was no difference in triglyceride concentration between these groups. Dogs on diet I for 12 to 16 months showed a further and substantial increase in plasma FA concentration; these changes were most marked in cholesteryl esters. Little or no lipoprotein with electrophoretic and ultracentrifugal properties of alpha-lipoprotein was present in the plasma. Immunotechniques showed that it was present. The composition of dietary FA had great influence in producing this hyperlipemia. Lipemia produced was not a simple reflection of the FA in these diets as evidenced by the increase in some FA, e.g., C_16∶1, which was absent in the experimental diets and C_18∶1, which contributed only 3.4% of the FA. Large increases in palmitoleate and oleate indicate synthesis or mobilization or both from other tissues. Diets composed predominantly of saturated medium chain length fatty acids, with or without added cholesterol were equally effective in the initiation of hyperlipemia. Data also suggest that added cholesterol is necessary for sustaining hyperlipemia. Presented in part at the AOCS Annual Meeting, New Orleans, April 1970.