Towards Application of One-Class Classification Methods to Medical Data

In the problem of one-class classification (OCC) one of the classes, the target class, has to be distinguished from all other possible objects, considered as nontargets. In many biomedical problems this situation arises, for example, in diagnosis, image based tumor recognition or analysis of electrocardiogram data. In this paper an approach to OCC based on a typicality test is experimentally compared with reference state-of-the-art OCC techniques-Gaussian, mixture of Gaussians, naive Parzen, Parzen, and support vector data description-using biomedical data sets. We evaluate the ability of the procedures using twelve experimental data sets with not necessarily continuous data. As there are few benchmark data sets for one-class classification, all data sets considered in the evaluation have multiple classes. Each class in turn is considered as the target class and the units in the other classes are considered as new units to be classified. The results of the comparison show the good performance of the typicality approach, which is available for high dimensional data; it is worth mentioning that it can be used for any kind of data (continuous, discrete, or nominal), whereas state-of-the-art approaches application is not straightforward when nominal variables are present.

Efeitos da exposição à nicotina e à fumaça de cigarro durante a lactação sobre o sistema de recompensa cerebral em ratos

Crianças de mães fumantes são mais suscetíveis a se tornarem adultos obesos e se viciarem em drogas ou alimentos palatáveis. Drogas e alimentos ativam a via mesolímbica de recompensa, causando sensação de prazer que induz ainda mais o consumo. Assim, avaliamos a relação entre a exposição apenas à nicotina ou à fumaça do cigarro durante a lactação com a preferência alimentar e sistema dopaminérgico de recompensa cerebral das proles, em dois modelos de programação: Modelo I: no 2o dia pós-natal (PN), lactantes receberam implante de minibombas osmóticas que liberam nicotina (NIC) ou salina (C), durante 14 dias. Em PN150 e novamente em PN160, as proles foram divididas em 4 grupos para um desafio alimentar: N-SC e C-SC que receberam ração padrão; N-SSD e C-SSD que podiam escolher livremente entre as dietas hiperlipídica e hiperglicídica. A ingestão alimentar foi avaliada após 12 h. As mães foram sacrificadas apenas na 21 da lactação (desmame) e as proles em PN15 (com nicotina), PN21 e PN170 (ausência da NIC). Ao desmame, as ratas lactantes NIC apresentaram menor conteúdo de tirosina hidroxilase (TH), maior OBRb e SOCS3 na area tegmentar ventral (VTA); menor TH, maior receptor de dopamina 1 (D1R), receptor de dopamina 2 (D2R) e transportador de dopamina (DAT) no núcleo accumbens (NAc); maior conteúdo de TH no estriado dorsal (DS); e maior D2R e SOCS3 no núcleo arqueado (ARC). Em PN15, os filhotes NIC apresentaram maior conteúdo de D1R, D2R e menor DAT no NAc, enquanto em PN21, apresentaram apenas menor DAT no DS, e menor conteúdo de pSTAT3 em ARC. Aos 170 dias, as proles SSD demonstraram maior preferência para a ração hiperlipídica. No entanto, os animais N-SSD consumiram mais ração hiperglicidica do que as proles C-SSD. A prole N apresentou menor conteúdo de D2R e DAT no NAc e menor D2R no ARC. Modelo II: as mães e suas proles foram divididas em: expostos à fumaça do cigarro (grupo S: 4 vezes / dia, do 3 ao 21 dia de lactação), e expostos ao ar filtrado (grupo C). Em PN175, as proles foram divididas em 4 grupos para o desafio alimentar S-SC, C-SC, S-SSD e C-SSD. A ingestão alimentar foi avaliada após 30 min e 12 h. Em PN180, as proles foram sacrificadas. O grupo S-SSD ingeriu mais das rações palatáveis do que o grupo C-SSD em 30 min e 12 h. Ambos os grupos preferiram a ração hiperlipídica. No entanto, os animais S-SSD consumiram mais ração hiperlipídica do que C-SSD em 30 min. A prole S apresentou menor conteúdo de TH no VTA, menor conteúdo de TH, D2R e maior conteúdo de D1R no NAc e menor OBRb no ARC. Demonstramos que tanto a nicotina isolada como a exposição à fumaça do cigarro durante a lactação resultaram em mudanças no sistema dopaminérgico das proles, programando o comportamento alimentar devido à diminuição da dopamina no NAc.

Condições de contorno tipo albedo para cálculos globais de reatores nucleares com o modelo de dois grupos de energia na formulação de transporte SN em geometria bidimensional cartesiana

Os eventos de fissão nuclear, resultados da interação dos nêutrons com os núcleos dos átomos do meio hospedeiro multiplicativo, não estão presentes em algumas regiões dos reatores nucleares, e.g., moderador, refletor, e meios estruturais. Nesses domínios espaciais não há geração de potência nuclear térmica e, além disso, comprometem a eficiência computacional dos cálculos globais de reatores nucleares. Propomos nesta tese uma estratégia visando a aumentar a eficiência computacional dessas simulações eliminando os cálculos numéricos explícitos no interior das regiões não-multiplicativas (baffle e refletor) em torno do núcleo ativo. Apresentamos algumas modelagens e discutimos a eficiência da aplicação dessas condições de contorno aproximadas tipo albedo para uma e duas regiões nãomultiplicativas, na formulação de ordenadas discretas (SN) para problemas de autovalor a dois grupos de energia em geometria bidimensional cartesiana. A denominação Albedo, palavra de origem latina para alvura, foi originalmente definida como a fração da luz incidente que é refletida difusamente por uma superfície. Esta denominação latina permaneceu como o termo científico usual em astronomia e, nesta tese, este conceito é estendido para reflexão de nêutrons. Estas condições de contorno tipo albedo SN não-convencional substituem aproximadamente as regiões de baffle e refletor no em torno do núcleo ativo do reator, desprezando os termos de fuga transversal no interior dessas regiões. Se o problema, em particular, não possui termos de fuga transversal, i.e., trata-se de um problema unidimensional, então as condições de contorno albedo, como propostas nesta tese, são exatas. Por eficiência computacional entende-se a análise da precisão dos resultados numéricos em comparação com o tempo de execução computacional de cada simulação de um dado problema-modelo. Resultados numéricos considerando dois problemas-modelo com de simetria são considerados para ilustrar esta análise de eficiência.

Discovering transcriptional modules by Bayesian data integration.

MOTIVATION: We present a method for directly inferring transcriptional modules (TMs) by integrating gene expression and transcription factor binding (ChIP-chip) data. Our model extends a hierarchical Dirichlet process mixture model to allow data fusion on a gene-by-gene basis. This encodes the intuition that co-expression and co-regulation are not necessarily equivalent and hence we do not expect all genes to group similarly in both datasets. In particular, it allows us to identify the subset of genes that share the same structure of transcriptional modules in both datasets. RESULTS: We find that by working on a gene-by-gene basis, our model is able to extract clusters with greater functional coherence than existing methods. By combining gene expression and transcription factor binding (ChIP-chip) data in this way, we are better able to determine the groups of genes that are most likely to represent underlying TMs. AVAILABILITY: If interested in the code for the work presented in this article, please contact the authors. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Effects of rotating inlet distortion on multistage compressor stability

In multi-spool engines, rotating stall in an upstream compressor will impose a rotating distortion on the downstream compressor, thereby affecting its stability margin. In this paper experiments are described in which this effect was simulated by a rotating screen upstream of several multistage low-speed compressors. The measurements are complemented by, and compared with, a theoretical model of multistage compressor response to speed and direction of rotation of an inlet distortion. For co-rotating distortions (i.e., distortions rotating in the same direction as rotor rotation), experiments show that the compressors exhibited significant loss in stability margin and that they could be divided into two groups according to their response. The first group exhibited a single peak in stall margin degradation when the distortion speed corresponded to roughly 50% of rotor speed. The second group showed two peaks in stall margin degradation corresponding to distortion speeds of approximately 25-35% and 70-75% of rotor speed. These new results demonstrate that multistage compressors can have more than a single resonant response. Detailed measurements suggest that the two types of behavior are linked to differences between the stall inception processes observed for the two groups of compressors and that a direct connection thus exists between the observed forced response and the unsteady flow phenomena at stall onset. For counter-rotational distortions, all the compressors tested showed minimal loss of stability margin. The results imply that counter-rotation of the fan and core compressor, or LP and HP compressors, could be a worthwhile design choice. Calculations based on the two-dimensional theoretical model show excellent agreement for the compressors which had a single peak for stall margin degradation. We take this first-of-a-kind comparison as showing that the model, though simplified, captures the essential fluid dynamic features of the phenomena. Agreement is not good for compressors which had two peaks in the curve of stall margin shift versus distortion rotation speed. The discrepancy is attributed to the three-dimensional and short length scale nature of the stall inception process in these machines; this includes phenomena that have not yet been addressed in any model.

Cluster analysis of heterogeneous rank data

Cluster analysis of ranking data, which occurs in consumer questionnaires, voting forms or other inquiries of preferences, attempts to identify typical groups of rank choices. Empirically measured rankings are often incomplete, i.e. different numbers of filled rank positions cause heterogeneity in the data. We propose a mixture approach for clustering of heterogeneous rank data. Rankings of different lengths can be described and compared by means of a single probabilistic model. A maximum entropy approach avoids hidden assumptions about missing rank positions. Parameter estimators and an efficient EM algorithm for unsupervised inference are derived for the ranking mixture model. Experiments on both synthetic data and real-world data demonstrate significantly improved parameter estimates on heterogeneous data when the incomplete rankings are included in the inference process.

Semi-supervised video segmentation using tree structured graphical models

We present a novel, implementation friendly and occlusion aware semi-supervised video segmentation algorithm using tree structured graphical models, which delivers pixel labels alongwith their uncertainty estimates. Our motivation to employ supervision is to tackle a task-specific segmentation problem where the semantic objects are pre-defined by the user. The video model we propose for this problem is based on a tree structured approximation of a patch based undirected mixture model, which includes a novel time-series and a soft label Random Forest classifier participating in a feedback mechanism. We demonstrate the efficacy of our model in cutting out foreground objects and multi-class segmentation problems in lengthy and complex road scene sequences. Our results have wide applicability, including harvesting labelled video data for training discriminative models, shape/pose/articulation learning and large scale statistical analysis to develop priors for video segmentation. © 2011 IEEE.

Design method of an innovative motor for an intra-aortic ventricular assist device

A permanent-magnet motor has been designed for an innovative axial-flow ventricular assist device (VAD), to be placed in the descending aorta, intended to offload the left ventricle and augment renal perfusion in patients with congestive heart failure (CHF). For this application, an intra-aortic impeller with a built-in permanent magnet rotor is driven by an extraaortic stator working in synchronism with the natural heart. To meet this need, a two-dimensional analytical model has been developed in the MATLAB environment to estimate machine parameters; finite element analysis (FEA) has been used to refine the results. A prototype blood pump equipped with an innovative motor designed from the procedure above has been tested in a mock loop representing the human circulatory system. The performance of VAD incorporating the motor is presented. © 2009 IEEE.

Aeroacoustic source measurement methods for characterizing the sound generated by ducted flow devices with higher-order modes.

The International Organization for Standardization (ISO) method 5136 is widely used in industry and academia to determine the sound power radiated into a duct by fans and other flow devices. The method involves placing the device at the center of a long cylindrical duct with anechoic terminations at each end to eliminate reflections. A single off-axis microphone is used on the inlet and outlet sides that can theoretically capture the plane-wave mode amplitudes but this does not provide enough information to fully account for higher-order modes. In this study, the "two-port" source model is formulated to include higher-order modes and applied for the first three modes. This requires six independent surface pressure measurements on each side or "port." The resulting experimental set-up is much shorter than the ISO rig and does not require anechoic terminations. An array of six external loudspeaker sources is used to characterize the passive part of the two-port model and the set-up provides a framework to account for transmission of higher-order modes through a fan. The relative importance of the higher-order modes has been considered and their effect on inaccuracies when using the ISO method to find source sound power has been analyzed.

Enlarging regions of stable running with segmented legs

In human and animal running spring-like leg behavior is found, and similar concepts have been demonstrated by various robotic systems in the past. In general, a spring-mass model provides self-stabilizing characteristics against external perturbations originated in leg-ground interactions and motor control. Although most of these systems made use of linear spring-like legs. The question addressed in this paper is the influence of leg segmentation (i.e. the use of rotational joint and two limb-segments) to the self-stability of running, as it appears to be a common design principle in nature. This paper shows that, with the leg segmentation, the system is able to perform self-stable running behavior in significantly broader ranges of running speed and control parameters (e.g. control of angle of attack at touchdown, and adjustment of spring stiffness) by exploiting a nonlinear relationship between leg force and leg compression. The concept is investigated by using a two-segment leg model and a robotic platform, which demonstrate the plausibility in the real world. ©2008 IEEE.

Confinement factor and absorption loss of AlInGaN based laser diodes emitting from ultraviolet to green

Confinement factor and absorption loss of AlInGaN based multiquantum well laser diodes (LDs) were investigated by numerical simulation based on a two-dimensional waveguide model. The simulation results indicate that an increased ridge height of the waveguide structure can enhance the lateral optical confinement and reduce the threshold current. For 405 nm violet LDs, the effects of p-AlGaN cladding layer composition and thickness on confinement factor and absorption loss were analyzed. The experimental results are in good agreement with the simulation analysis. Compared to violet LD, the confinement factors of 450 nm blue LD and 530 nm green LD were much lower. Using InGaN as waveguide layers that has higher refractive index than GaN will effectively enhance the optical confinement for blue and green LDs. The LDs based on nonpolar substrate allow for thick well layers and will increase the confinement factor several times. Furthermore, the confinement factor is less sensitive to alloys composition of waveguide and cladding layers, being an advantage especially important for ultraviolet and green LDs.

Intraband relaxation and its influences on quantum dot lasers

A comprehensive two-level numerical model is developed to describe carrier distribution in a quantum-dot laser. Light-emission spectra with different intraband relaxation rates (2ps, 7.5ps and 20ps) are calculated and analysed to investigate the influence of relaxation rates on performance of the quantum-dot laser. The results indicate that fast intraband relaxation favours not only the ground state single mode operation but also the higher injection efficiency.

Anti-Stokes photoluminescence in ZnO microcrystal

Low temperature (10 K) strong anti-Stokes photoluminescence (ASPL) of ZnO microcrystal excited by low power cw 532 nm laser is reported here. Energy upconversion of 1.1 eV is obtained in our experiment with no conventional nonlinear effect. Through the study of the normal photoluminescence and temperature dependence of ASPL we conclude that the green band luminescence in ZnO is related to deep donor to valance band transition. Using the two-step two-photon absorption model, we provide a plausible mechanism leading to the ASPL phenomenon in our experiment. (c) 2006 American Institute of Physics.

Electroluminescence from Au/(SiO2/Si/SiO2) nanometer double barrier/p-Si structures and its mechanism

SiO2/Si/SiO2 nanometer double barriers (SSSNDB) with Si layers of twenty-seven different thicknesses in a range of 1-5 nm with an interval of 0.2 nm have been deposited on p-Si substrates using two-target alternative magnetron sputtering. Electroluminescence (EL) from the semitransparent Au film/SSSNDB/p-Si diodes and from a control diode without any Si layer have been observed under forward bias. Each EL spectrum of all these diodes can be fitted by two Gaussian bands with peak energies of 1.82 and 2.25 eV, and full widths at half maximum of 0.38 and 0.69 eV, respectively. It is found that the current, EL peak wavelength and intensities of the two Gaussian bands of the Au/SSSNDB/p-Si structure oscillate synchronously with increasing Si layer thickness with a period corresponding to half a de Broglie wavelength of the carriers. The experimental results strongly indicate that the EL originates mainly from two types of luminescence centres with energies of 1.82 and 2.25 eV in the SiO2 barriers, rather than from the nanometer Si well in the SSSNDB. The EL mechanism is discussed in detail.

Annealing behaviors of photoluminescence from SiOx : H

The strong photoluminescence (PL) of SiOx:H prepared by plasma enhanced chemical vapor deposition has been systematically studied in conjunction with infrared and micro-Raman spectra. We have found that each PL spectrum is comprised of two Gaussian components, a main band and a shoulder. The main band might originate from amorphous silicon clusters embedded in die SiOx network, and its redshift with annealing temperature is due to expansion of the silicon clusters. The shoulder remains at about 835 nm in spite of the annealing temperature and possibly comes from luminescent defect centers. The enhanced PL spectra after 1170 degrees C annealing are attributed to the quantum confinement effects of nanocrystalline silicon embedded in the SiO2 matrix. (C) 1998 American Institute of Physics.