Modeling and Classifying Breast Tissue Density in Mammograms

We present a new approach to model and classify breast parenchymal tissue. Given a mammogram, first, we will discover the distribution of the different tissue densities in an unsupervised manner, and second, we will use this tissue distribution to perform the classification. We achieve this using a classifier based on local descriptors and probabilistic Latent Semantic Analysis (pLSA), a generative model from the statistical text literature. We studied the influence of different descriptors like texture and SIFT features at the classification stage showing that textons outperform SIFT in all cases. Moreover we demonstrate that pLSA automatically extracts meaningful latent aspects generating a compact tissue representation based on their densities, useful for discriminating on mammogram classification. We show the results of tissue classification over the MIAS and DDSM datasets. We compare our method with approaches that classified these same datasets showing a better performance of our proposal

Prediction of individual automobile RBNS claim reserves in the context of Solvency II

Automobile bodily injury (BI) claims remain unsettled for a long time after the accident. The estimation of an accurate reserve for Reported But Not Settled (RBNS) claims is therefore vital for insurers. In accordance with the recommendation included in the Solvency II project (CEIOPS, 2007) a statistical model is here implemented for RBNS reserve estimation. Lognormality on empirical compensation cost data is observed for different levels of BI severity. The individual claim provision is estimated by allocating the expected mean compensation for the predicted severity of the victim’s injury, for which the upper bound is also computed. The BI severity is predicted by means of a heteroscedastic multiple choice model, because empirical evidence has found that the variability in the latent severity of injured individuals travelling by car is not constant. It is shown that this methodology can improve the accuracy of RBNS reserve estimation at all stages, as compared to the subjective assessment that has traditionally been made by practitioners.

Modelo de calibração da concentração de metilmetacrilato em solução aquosa utilizando espectroscopia de absorção no ultravioleta

Dilutions of methylmetacrylate ranging between 1 and 50 ppm were obtained from a stock solution of 1 ml of monomer in 100 ml of deionised water, and were analyzed by an absorption spectrophotometer in the UV-visible. Absorbance values were used to develop a calibration model based on the PLS, with the aim to determine new sample concentrations. The number of latent variables used was 6, with the standard errors of calibration and prediction found to be 0,048 ml/100 ml and 0,058 ml/100 ml. The calibration model was successfully used to calculate the concentration of monomer released in water, where complete dentures were kept for one hour after polymerization.

A state-space approach for measuring regional manufacturing production indices

In this paper we propose a latent variable model, in the spirit of Israilevich and Kuttner (1993), to measure regional manufacturing production. To test the validity of the proposed methodology, we have applied it for those Spanish regions that have a direct quantitative index. The results demonstrate the accuracy of the methodology proposed and show that it can overcome some of the difficulties of the indirect method applied by the INE, the Spanish National Institute of Statistics.

Advanced chemometric methods: Applicability on industrial data

Recent years have produced great advances in the instrumentation technology. The amount of available data has been increasing due to the simplicity, speed and accuracy of current spectroscopic instruments. Most of these data are, however, meaningless without a proper analysis. This has been one of the reasons for the overgrowing success of multivariate handling of such data. Industrial data is commonly not designed data; in other words, there is no exact experimental design, but rather the data have been collected as a routine procedure during an industrial process. This makes certain demands on the multivariate modeling, as the selection of samples and variables can have an enormous effect. Common approaches in the modeling of industrial data are PCA (principal component analysis) and PLS (projection to latent structures or partial least squares) but there are also other methods that should be considered. The more advanced methods include multi block modeling and nonlinear modeling. In this thesis it is shown that the results of data analysis vary according to the modeling approach used, thus making the selection of the modeling approach dependent on the purpose of the model. If the model is intended to provide accurate predictions, the approach should be different than in the case where the purpose of modeling is mostly to obtain information about the variables and the process. For industrial applicability it is essential that the methods are robust and sufficiently simple to apply. In this way the methods and the results can be compared and an approach selected that is suitable for the intended purpose. Differences in data analysis methods are compared with data from different fields of industry in this thesis. In the first two papers, the multi block method is considered for data originating from the oil and fertilizer industries. The results are compared to those from PLS and priority PLS. The third paper considers applicability of multivariate models to process control for a reactive crystallization process. In the fourth paper, nonlinear modeling is examined with a data set from the oil industry. The response has a nonlinear relation to the descriptor matrix, and the results are compared between linear modeling, polynomial PLS and nonlinear modeling using nonlinear score vectors.

Determinação de sibutramina em formas farmacêuticas através de espectroscopia no infravermelho com refletância difusa e métodos de calibração multivariada

The goal of this work is the development and validation of an analytical method for fast quantification of sibutramine in pharmaceutical formulations, using diffuse reflectance infrared spectroscopy and partial least square regression. The multivariate model was elaborated from 22 mixtures containing sibutramine and excipients (lactose, microcrystalline cellulose, colloidal silicon dioxide and magnesium stearate) and using fragmented (750-1150/ 1350-1500/ 1850-1950/ 2600-2900 cm-1) and smoothing spectral data. Using 10 latent variables, excellent predictive capacity were observed in the calibration (n=20, RMSEC=0.004, R= 0.999) and external validation (n=5, RMSEC= 9.36, R=0.999) phases. In the analysis of synthetic mixtures the precision (SD=3,47%) was compatible with the rules of the Agencia Nacional de Vigilância Sanitária (ANVISA-Brazil). In the analysis of commercial drugs good agreement was observed between spectroscopic and chromatographic methods.

Caracterização espectroscópica multivariada do potencial antioxidante de vinhos

In this work the antioxidant capacity of red wine samples was characterized by conventional spectroscopic and chromatographic methodologies, regarding chemical parameters like color, total polyphenolic and resveratrol content, and antioxidant activity. Additionally, multivariate calibration models were developed to predict the antioxidant activity, using partial least square regression and the spectral data registered between 400 and 800 nm. Even when a close correlation between the evaluated parameters has been expected many inconsistencies were observed, probably on account of the low selectivity of the conventional methodologies. Models developed from mean-centered spectra and using 4 latent variables allowed high prevision capacity of the antioxidant activity, permitting relative errors lower than 3%.

Determinação simultânea de ácido kójico e hidroquinona por espectrofotometria visível e calibração multivariada

A multivariate spectrophotometric method was developed for analysis of kojic acid/hydroquinone associations in skin whitening cosmetics. The method is based on the reaction between kojic acid and Fe3+ and on the reduction of Fe3+ by hydroquinone and further complexation of Fe2+ with 1,10-phenanthroline. The multivariate model was developed by Partial Least Squares Regression (PLSR), using 25 synthetic mixtures and mean-centered spectral data (350-380 nm). The use of 3 (kojic acid) and 2 (hydroquinone) latent variables permits the observation of mean errors of about 5% in the external validation phase.

Monitorando atividades enzimáticas com sondas fluorogênicas

Latent fluorogenic probes are essential tools for molecular and chemical biology, providing valuable information about enzymatic activity and occurrence. In this review, a brief outline of fluorophores and latent fluorogenic probes is given. Furthermore, advances and challenges in the development of fluorogenic chemical probes to visualize enzymatic activities (hydrolases and oxidoreductases) of biotechnological and biomedical interest are highlighted, including some methodologies for intracellular imaging.

Garlic viral complex: identification of Potyviruses and Carlavirus in Central Brazil

Garlic viruses often occur in complex infections in nature. In this study, a garlic virus complex, collected in fields in Brazil, was purified. RT-PCR was performed using specific primers designed from the consensus regions of the coat protein genes of Onion yellow dwarf virus, a garlic strain (OYDV-G) and Leek yellow stripe virus (LYSV). cDNA of Garlic common latent virus (GCLV) was synthesized using oligo-dT and random primers. By these procedures individual garlic virus genomes were isolated and sequenced. The nucleotide sequence analysis associated with serological data reveals the presence of two Potyvirus OYDV-G and LYSV, and GCLV, a Carlavirus, simultaneously infecting garlic plants. Deduced amino acid sequences of the Brazilian isolates were compared with related viruses reported in different geographical regions of the world. The analysis showed closed relations considering the Brazilian isolates of OYDV-G and GCLV, and large divergence considering LYSV isolate. The detection of these virus species was confirmed by specific reactions observed when coat protein genes of the Brazilian isolates were used as probes in dot-blot and Southern blot hybridization assays. In field natural viral re-infection of virus-free garlic was evaluated.

Sequence analysis of the capsid protein gene of an isolate of Apple stem grooving virus, and its survey in Southern Brazil

Apple stem grooving virus (ASGV) is one of the most important viruses infecting fruit trees. This study aimed at the molecular characterization of ASGV infecting apple (Malus domestica) plants in Santa Catarina (SC). RNA extracted from plants infected with isolate UV01 was used as a template for RT-PCR using specific primers. An amplified DNA fragment of 755 bp was sequenced. The coat protein gene of ASGV isolate UV01 contains 714 nucleotides, coding for a protein of 237 amino acids with a predicted Mr of approximately 27 kDa. The nucleotide and the deduced amino acid sequences of the coat protein gene showed identities of 90.9% and 97.9%, respectively, with a Japanese isolate of ASGV. Very high amino acid homologies (98.7%) were also found with Citrus tatter leaf capillovirus (CTLV), a very close relative of ASGV. These results indicate low coat protein gene variability among Capillovirus isolates from distinct regions. In a restricted survey, mother stocks in orchards and plants introduced into the country for large scale fruit production were indexed and shown to be infected by ASGV (20%), usually in a complex with other (latent) apple viruses (80%).

Detecção de um isolado de Grapevine virus A e caracterização do gene da proteína capsidial

O Grapevine virus A (GVA) está associado à "Acanaladura do lenho de Kober", uma doença do complexo rugoso da videira (Vitis spp.). Neste trabalho, um isolado brasileiro de GVA (GVA-RS) foi caracterizado biologicamente por transmissão mecânica para cinco hospedeiras herbáceas e por enxertia na videira indicadora cv. Kober 5BB, e também por sorologia. O RNA total foi extraído de videira infetada cv. Pirovano 65. Para a RT-PCR, dois pares de oligonucleotídeos foram utilizados. Dois fragmentos de DNA, 430 e 451 pb, apresentando sobreposição parcial de nucleotídeos, foram amplificados por PCR. A seqüência do gene da proteína capsidial do GVA-RS com 597 nucleotídeos e 198 aminoácidos deduzidos, com massa molecular calculada de 21,6 kDa, foi alinhada a outros isolados virais. As seqüências de nucleotídeos e aminoácidos deduzidos do GVA-RS apresentaram maior identidade, 91,4% e 95,4%, respectivamente, com um isolado italiano. O GVA-RS apresentou expressiva divergência dos Vitivirus Heracleum latent virus (HLV), Grapevine virus B (GVB) e Grapevine virus D (GVD), com identidade de nucleotídeos variando de 76% a 83,1%.

Polyclonal antibodies to the coat protein of Apple stem grooving virus expressed in Escherichia coli: production and use in immunodiagnosis

The coat protein gene of Apple stem grooving virus (ASGV) was amplified by RT-PCR, cloned, sequenced and subcloned in the expression vector pMal-c2. This plasmid was used to transform Escherichia coli BL21c+ competent cells. The ASGV coat protein (cp) was expressed as a fusion protein containing a fragment of E. coli maltose binding protein (MBP). Bacterial cells were disrupted by sonication and the ASGVcp/MBP fusion protein was purified by amylose resin affinity chromatography. Polyclonal antibodies from rabbits immunized with the fusion protein gave specific reactions to ASGV from infected apple (Malus domestica) cv. Fuji Irradiada and Chenopodium quinoa at dilutions of up to 1:1,000 and 1:2,000, respectively, in plate trapped ELISA. The ASGVcp/MBP fusion protein reacted to a commercial antiserum against ASGV in immunoblotting assay. The IgG against ASGVcp/MBP performed favorably in specificity and sensitivity to the virus. This method represents an additional tool for the efficient ASGV-indexing of apple propagative and mother stock materials, and for use in support of biological and molecular techniques.

Epidemiological analysis of clones and cultivars of potato in soil naturally infested with Ralstonia solanacearum biovar 2

The objectives of this study were to evaluate the progress of Ralstonia solanacearum bacterial potato wilt biovar 2 (race 3) in 14 potato (Solanum tuberosum L.) cultivars or clones, the resistance of potato clone MB 03 (selected in Brasília, Brazil) to race 1 of R. solanacearum, and the occurrence of the pathogen in tubers harvested from asymptomatic potato plants. During the spring (September to the end of November in the southern hemisphere) of 1999 and 2000, 14 cultivars or clones were grown in a field naturally infested with R. solanacearum biovar 2, in Caxias do Sul, RS. The number of wilted potato plants was recorded each week and a disease progress curve plotted, the resistance of the potato genotypes to bacterial wilt being evaluated by determining the area under the curve. Various models were evaluated to fit the curves, with the logistic model being the best fit. At the end of each growing season tubers produced by asymptomatic plants were harvested and stored until budding and then tested for the presence of R. solanacearum. Cultivar Cruza 148 and clone MB 03 were the most resistant but both showed tubers with latent infections. The epidemiological implications of the incidence of R. solanacearum biovar 2 (race 3) in potato crops, as well as the resistance of certain genotypes that may harbor latent infections, are important aspects to be considered in the integrated management of bacterial wilt.

Effect of temperature, wetness duration and cultivar on the development of anthracnose in guava fruits

The development of a large number of postharvest diseases is closely associated with fruit ripeness. Environmental conditions may affect both the pathogen development and the fruit ripening rate. The aim of this study was to determine the most favorable temperature and wetness duration to the development of anthracnose in guava fruits. Cultivars 'Kumagai' (white pulp) and 'Pedro Sato' (red pulp) were inoculated with a conidial suspension of Colletotrichum gloeosporioides and C. acutatum and incubated at constant temperature ranging from 10 to 35ºC and wetness duration of 6 and 24 hours. Disease severity and incidence were evaluated at every two days during 12 days. No infection occurred at 10 and 35ºC, regardless of the wetness duration. The optimum conditions for fruit infection were 26 and 27ºC for 'Kumagai' and 25 and 26ºC for 'Pedro Sato', adopting 24 hours of wetness. In general, the disease development in 'Kumagai' cultivar was more affected by the wetness period, compared to 'Pedro Sato'. Disease severity for 'Kumagai' fruits was maximal between 25 and 30ºC , depending on the Colletotrichum species. Regarding 'Pedro Sato', the mean diameter of lesions was greater in fruits stored at 20, 25 and 30ºC , compared to 'Kumagai' cultivar, depending on the wetness period and the species. The incubation period (between 6 and 7 days) and the latent period (between 8 and 10 days) were minimal at 30ºC. The data generated in this study will be useful either for the development of a disease warning system or for the increase in the shelf life of guavas in the postharvest.