Comparing Prediction of Early TBI Mortality with Multilayer Perceptron Neural Network and Convolutional Neural Network.

In this work, we compare the performance of a multilayer perceptron neural network and convolutional networks for the prediction of 14-day mortality in patients with TBI, using a database obtained in a low-and middle-income country, with 529 records and 16 predictor variables. The missing values of several variables were filled in with techniques such as decision tree, random forest, k-nearest-neighbor and linear regression. In the simulation of neural networks, several optimization methods were used, such as RMSProp, Adam, Adamax and SGDM. The best results obtained for the prediction rate were an accuracy of 0.845 and an area under the ROC curve of 0.911. Clinical Relevance- This proposes the prediction of early mortality in patients with TBI with an area under ROC curve of 0.911.

A New Approach to Classify Cardiac Arrythmias Using 2D Convolutional Neural Networks.

Cardiovascular diseases are the number one cause of death worldwide. Detecting cardiovascular diseases in its early stages could effectively reduce the mortality rate by providing timely treatment. In this study, we propose a new methodology to detect arrythmias, using 2D Convolutional Neural Networks. The main characteristic of the proposed methodology is the use of 15 x15 pixels gray-level images, containing the values of a heartbeat of the ECG signal. This work aims to detect 17 arrythmias. To validate and test the proposed methodology, MIT-BIH database, the main benchmark database available in literature, was used. When compared to other results previously published, the obtained precision, 92.31%, is in the state-of-the-art.Clinical Relevance- The presented work provides an automatic method to detect arrythmias in ECG signals by a new methodology.

Lung Region Segmentation in Chest X-Ray Images using Deep Convolutional Neural Networks.

Lung cancer is, by far, the leading cause of cancer death in the world. Tools for automated medical imaging analysis development of a Computer-Aided Diagnosis method comprises several tasks. In general, the first one is the segmentation of region of interest, for example, lung region segmentation from Chest X-ray imaging in the task of detecting lung cancer. Deep Convolutional Neural Networks (DCNN) have shown promising results in the task of segmentation in medical images. In this paper, to implement the lung region segmentation task on chest X-ray images, was evaluated three different DCNN architectures in association with different regularization (Dropout, L2, and Dropout + L2) and optimization methods (SGDM, RMSPROP and ADAM). All networks were applied in the Japanese Society of Radiological Technology (JSRT) database. The best results were obtained using Dropout + L2 as regularization method and ADAM as optimization method. Considering the Jaccard Coefficient obtained (0.97967 ± 0.00232) the proposal outperforms the state of the art.Clinical Relevance- The presented method reduces the time that a professional takes to perform lung segmentation, improving the effectiveness.

Automatic Bacillus Detection in Light Field Microscopy Images Using Convolutional Neural Networks and Mosaic Imaging Approach.

Tuberculosis (TB) is one of the top 10 causes of death worldwide. The diagnosis and treatment of TB in its early stages is fundamental to reducing the rate of people affected by this disease. In order to assist specialists in the diagnosis in bright field smear images, many studies have been developed for the automatic Mycobacterium tuberculosis detection, the causative agent of Tb. To contribute to this theme, a method to bacilli detection associating convolutional neural network (CNN) and a mosaic-image approach was implemented. The propose was evaluated using a robust image dataset validated by three specialists. Three CNN architectures and 3 optimization methods in each architecture were evaluated. The deeper architecture presented better results, reaching accuracies values above 99%. Other metrics like precision, sensitivity, specificity and F1-score were also used to assess the CNN models performance.

Supplementation of lamb ewes with different protein sources in deferred marandu palisadegrass (Brachiaria brizantha cv. marandu) pasture / [Suplementação de matrizes ovinas com diferentes fontes de proteína em pasto diferido de capim-marandu (Brachiaria brizantha cv. Marandu)]

The objective of this study was to evaluate the effect of different protein sources on the performance of lamb ewes and their offspring supplemented in deferred pasture. The treatments were the different protein sources: soybean meal, cotton cake, babassu cake and coconut cake. The ewes were weighed and evaluated for their body condition score during the gestation, parturition and weaning phases. The lambs were weighed at birth and at weaning. The structural variables and chemical composition of the pasture were evaluated during the usage period. Grass height and forage mass were higher when the ewes were in the gestation and weaning stages. The leaf mass was not altered during the pasture usage with a mean value of 566.1kg ha-1 DM. The highest levels of crude protein in the leaf blade were observed during the parturition and weaning phases. The body weight of the ewes did not differ between the sources with averages 45.0, 39.9 and 36.6kg for the gestation, parturition and weaning phases, respectively. Weights at birth and at weaning had an effect from protein sources, with the highest values observed in the lambs of ewes who received soybean meal and babassu cake supplementation. The lambs of ewes supplemented with soybean had higher average daily gain and total gain. Babassu, coconut and cotton cakes can replace soybean meal in supplementation of ewes in deferred pasture.(AU)

O objetivo deste estudo foi avaliar o efeito de diferentes fontes de proteína no desempenho de ovelhas e suas crias suplementadas em pastagem diferida. Os tratamentos foram as diferentes fontes proteicas: farelo de soja, tortas de algodão, babaçu e coco. As ovelhas foram pesadas e avaliadas quanto ao escore de condição corporal durante as fases de gestação, parição e desmame. Os cordeiros foram pesados no nascimento e no desmame. As variáveis estruturais e a composição química do pasto foram avaliadas durante o período de uso. A altura do dossel e a massa de forragem foram maiores quando as ovelhas estavam nas fases de gestação e desmame. A massa foliar não foi alterada durante o uso do pasto, com um valor médio de 566,1kg ha-1 MS. Os maiores teores de proteína bruta na lâmina foliar foram observados durante as fases de parto e desmame. O peso corporal das ovelhas não diferiu entre as fontes proteicas, com médias de 45,0, 39,9 e 36,6kg para as fases de gestação, parto e desmame, respectivamente. Os pesos ao nascimento e ao desmame tiveram efeito das fontes proteicas, com os maiores valores observados nos cordeiros de ovelhas que receberam suplementação com farelo de soja e torta de babaçu. Os cordeiros suplementados com soja apresentaram maior ganho médio diário e ganho total. As tortas de babaçu, coco e algodão podem substituir o farelo de soja na suplementação de ovelhas em pastagem diferida.(AU)

Using Convolutional Neural Networks for Classification of Bifurcation Regions in IVOCT Images.

Optical Coherence Tomography (OCT) technology enabled the experts to analyze coronary lesions from high-resolution intravascular images. Studies have shown the relationship between bifurcation regions and a higher occurrence of wall thickening and lesions in these areas. Some level of automation could benefit experts, since examining pullback frames is a laborious and time-consuming task. Although Convolutional Neural Networks (CNN) have shown promising results in classification tasks of medical images, we did not identify the use of CNN's in IVOCT images to classify bifurcation regions in the literature. In this work, we evaluated a CNN architecture in the bifurcation classification task trained with IVOCT images from 9 pullbacks from 9 different patients. We used data augmentation to balance the dataset, due to the low amount of bifurcation-labeled frames. Our classification results are comparable to other works in the literature, presenting better result in AUC (99.70%).

Lumen Segmentation in Optical Coherence Tomography Images using Convolutional Neural Network.

Lumen segmentation in Optical Coherence Tomography (OCT) images is a very important step to analyze points of interest that may help on atherosclerosis diagnostic and treatment. Past studies use many different methods to segment the lumen in IVOCT images, like level set, morphological reconstruction, Markov random fields, and Otsu binarization. Despite Convolutional Neural Networks (CNN) have shown promising results in the image processing area, we did not identify, in the literature, works applying CNN in IVOCT images. In this paper, we present the lumen segmentation using CNN. We evaluated three different CNN architectures. The CNNs were evaluated using three versions from the image dataset, differing from each other by image size (768x768 pixels and 192x192 pixels), and by coordinate system representation (Cartesian and polar). The best results, Accuracy, Dice index and Jaccard index of over 99%, 98% and 97%, respectively, were obtained with the smallest size images represented by polar coordinate system.

Computational engineering of cellulase Cel9A-68 functional motions through mutations in its linker region.

Microbial cellulosic degradation by cellulases has become a complementary approach for biofuel production. However, its efficiency is hindered by the recalcitrance of cellulose fibres. In this context, computational protein design methods may offer an efficient way to obtain variants with improved enzymatic activity. Cel9A-68 is a cellulase from Thermobifida fusca that is still active at high temperatures. In a previous work, we described a collective bending motion, which governs the overall cellulase dynamics. This movement promotes the approximation of its CBM and CD structural domains (that are connected by a flexible linker). We have identified two residues (G460 and P461) located at the linker that act as a hinge point. Herein, we applied a new level of protein design, focusing on the modulation of this collective motion to obtain cellulase variants with enhanced functional dynamics. We probed whether specific linker mutations would affect Cel9A-68 dynamics through computational simulations. We assumed that P461G and G460+ (with an extra glycine) constructs would present enhanced interdomain motions, while the G460P mutant would be rigid. From our results, the P461G mutation resulted in a broader exploration of the conformational space, as confirmed by clustering and free energy analyses. The WT enzyme was the most rigid system. However, G460P and P460+ explored distinct conformational states described by opposite directions of low-frequency normal modes; they sampled preferentially closed and open conformations, respectively. Overall, we highlight two significant findings: (i) all mutants explored larger conformational spaces than the WT; (ii) the selection of distinct conformational populations was intimately associated with the mutation considered. Thus, the engineering of Cel9A-68 motions through linker mutations may constitute an efficient way to improve cellulase activity, facilitating the disruption of cellulose fibres.

Effect of overexpression of citrus 9-cis-epoxycarotenoid dioxygenase 3 (CsNCED3) on the physiological response to drought stress in transgenic tobacco.

9-cis-epoxycarotenoid dioxygenase (NCED) encodes a key enzyme in abscisic acid (ABA) biosynthesis. Little is known regarding the regulation of stress response by NCEDs at physiological levels. In the present study, we generated transgenic tobacco overexpressing an NCED3 ortholog from citrus (CsNCED3) and investigated its relevance in the regulation of drought stress tolerance. Wild-type (WT) and transgenic plants were grown under greenhouse conditions and subjected to drought stress for 10 days. Leaf predawn water potential (Ψwleaf), stomatal conductance (gs), net photosynthetic rate (A), transpiration rate (E), instantaneous (A/E) and intrinsic (A/gs) water use efficiency (WUE), and in situ hydrogen peroxide (H2O2) and abscisic acid (ABA) production were determined in leaves of irrigated and drought-stressed plants. The Ψwleaf decreased throughout the drought stress period in both WT and transgenic plants, but was restored after re-watering. No significant differences were observed in gs between WT and transgenic plants under normal conditions. However, the transgenic plants showed a decreased (P ≤ 0.01) gs on the 4th day of drought stress, which remained lower (P ≤ 0.001) than the WT until the end of the drought stress. The A and E levels in the transgenic plants were similar to those in WT; therefore, they exhibited increased A/gs under drought conditions. No significant differences in A, E, and gs values were observed between the WT and transgenic plants after re-watering. The transgenic plants had lower H2O2 and higher ABA than the WT under drought conditions. Our results support the involvement of CsNCED3 in drought avoidance.

A new method for recognizing hand configurations of Brazilian gesture language.

This paper describes a new method for recognizing hand configurations of the Brazilian Gesture Language - LIBRAS - using depth maps obtained with a Kinect® camera. The proposed method comprised three phases: hand segmentation, feature extraction, and classification. The segmentation phase is independent from the background and depends only on pixel depth information. Using geometric operations and numerical normalization, the feature extraction process was done independent from rotation and translation. The features are extracted employing two techniques: (2D)2LDA and (2D)2PCA. The classification is made with a novelty classifier. A robust database was constructed for classifier evaluation, with 12,200 images of LIBRAS and 200 gestures of each hand configuration. The best accuracy obtained was 95.41%, which was greater than previous values obtained in the literature.

Breast tumor classification in ultrasound images using neural networks with improved generalization methods.

Mammography, scintimammography and ultrasound images have been used to increase the specificity of breast cancer image diagnosis. Concerning breast cancer image diagnosis with ultrasound, some results found in the literature show better performance of morphological features in breast cancer lesion differentiation and that a reduced set of features shows a better performance than a large set of features. In this study we evaluated the performance of neural network classifiers, with different training stop criteria: mean square error, early stop and regularization. The last two criteria were developed to improve neural network generalization. Different sets of morphological features were used as neural network inputs. Training sets comprised of 22, 8, 7, 6, 5 and 4 features were employed. To select reduced sets of features, a scalar selection technique with correlation was used. The best results obtained for accuracy and area under the ROC curve were 96.98% and 0.98, respectively. The performance obtained with all 22 features is slightly better than the one obtained with a reduced set of features.

TcCYS4, a cystatin from cocoa, reduces necrosis triggered by MpNEP2 in tobacco plants.

In Brazil, most cocoa bean production occurs in Southern Bahia. Witches' broom disease arrived in this area in 1989 and has since caused heavy losses in production. The disease is caused by the basidiomycete fungus Moniliophthora perniciosa, a hemibiotrophic fungus that produces the necrosis and ethylene-inducting protein (MpNEP2) during infection; this protein can activate cysteine proteases and induce programmed cell death. Cysteine proteases can be modulated by cystatin. In this study, we overexpressed TcCYS4, a cocoa cystatin, in tobacco plants and evaluated the effect on MpNEP2 in model plants. Tccys4 cDNA was cloned into the pCAMBIA 1390 vector and inserted into the tobacco plants via Agrobacterium tumefaciens. Transgene expression was analyzed by reverse transcription-quantitative PCR and Western blot analysis. Transcript and protein levels in Tcccys4:tobacco lines were 8.9- and 1.5-fold higher than in wild-type plants (wt). Tcccys4:tobacco lines showed no change in growth compared to wt plants. CO2 net assimilation (A) increased in Tcccys4:tobacco lines compared to wt plants. Only one line showed statistically significant stomatal conductance (gs) and transpiration rate (E) changes. MpNEP2 was infiltered into the foliar mesophyll of Tcccys4:tobacco lines and wt plants, and necrotic lesions were attenuated in lines highly expressing Tccys4. Our results suggest that cocoa cystatin TcCYS4 affects MpNEP2 activity related to the progression of programmed cell death in tobacco plants. This may occur through the action of cystatin to inhibit cysteine proteases activated by MpNEP2 in plant tissues. Further studies are necessary to examine cystatin in the Theobroma cacao-M. perniciosa pathosystem.

Evaluation of the reasons for nonacceptance of kidneys retrieved or offered in Rio Grande do Sul and Pais Vasco.

In Rio Grande do Sul (RS), as in Pais Vasco (PV), some kidneys are retrieved or offered and not accepted for transplantation. This study aimed to evaluate the profile of the available kidneys and the reasons for them not being accepted in the 2 regions, and to compare the characteristics of the organs and reasons for refusal. All of the kidneys retrieved or offered in RS in December 2012 and in PV from September to December 2012 were evaluated. Data were collected from each local donation registry. There were 61 kidneys available in RS and 61 in PV in the study period. Of these, 16 kidneys (26%) in RS and 27 (44%) in PV were not implanted. The age of the donors was higher in PV (59 years) than in RS (45 years; P = .000), as was the age of the donors of accepted kidneys (62 and 41 years old, respectively; P = .000). The proportion of donors considered to be "extended criteria" was higher in PV (78%) than in RS (47%; P = .001), and the refusal rate of the kidneys from these donors was the same in the 2 regions. The reasons for not using the kidneys in RS and in PV were similar and absolute. It is concluded that there is no organ waste in the 2 regions, but that the offer of kidneys can be expanded in RS by considering elderly donors for evaluation, even if this means a higher number of refused organs.

Phosphate-induced-1 gene from Eucalyptus (EgPHI-1) enhances osmotic stress tolerance in transgenic tobacco.

Environmental stresses such as drought, freezing, and high salinity induce osmotic stress in plant cells. The plant response to osmotic stress involves a number of physiological and developmental changes, which are made possible, in part, by the modulation of the expression of specific genes. Phosphate-induced-1 gene (PHI-1) was first isolated from phosphate-treated phosphate-starved tobacco cell cultures as a stress-inducible gene, which is presumably related to intracellular pH maintenance; however, the role of the PHI-1 gene product has not yet been clarified. A gene encoding a predicted protein with high similarity to tobacco PHI-1, named EgPHI-1, was previously identified in Eucalyptus by comparative transcriptome analysis of xylem cells from species of contrasting phenotypes for wood quality and growth traits. Here, we show that the overexpression of EgPHI-1 in transgenic tobacco enhances tolerance to osmotic stress. In comparison with wild-type plants, EgPHI-1 transgenic plants showed a significant increase in root length and biomass dry weight under NaCl-, polyethylene glycol, and mannitol-induced osmotic stresses. The enhanced stress tolerance of transgenic plants was correlated with increased endogenous protein levels of the molecular chaperone binding protein BiP, which in turn was correlated with the EgPHI-1 expression level in the different transgenic lines. These results provide evidence about the involvement of EgPHI-1 in osmotic stress tolerance via modulation of BiP expression, and pave the way for its future use as a candidate gene for engineering tolerance to environmental stresses in crop plants.

A sputum smear microscopy image database for automatic bacilli detection in conventional microscopy.

In this work, we present an image database for automatic bacilli detection in sputum smear microscopy. The database comprises two parts. The first one, called the autofocus database, contains 1200 images with resolution of 2816 × 2112 pixels. This database was obtained from 12 slides, with 10 fields per slide. Each stack is composed of 10 images, with the fifth image in focus. The second one, called the segmentation and classification database, contains 120 images with resolution of 2816×2112 pixels. This database was obtained from 12 slices, with 10 fields per slice. In both databases, the images were acquired from fields of slides stained with the standard Kinyoun method. In both databases, accordingly to the background content, the images were classified as belonging to high background content or low background content. In all 120 images of segmentation and classification database, the identified objects were enclosed within a geometric shape by a trained technician. A true bacillus was enclosed in a circle. An agglomerated bacillus was enclosed by a rectangle and a doubtful bacillus (the image focus or geometry does not allow a clear identification of the object) was enclosed by a polygon. These marked objects could be used as a gold standard to calculate the accuracy, sensitivity and specificity of bacilli recognition.

Proteomic response of Moniliophthora perniciosa exposed to pathogenesis-related protein-10 from Theobroma cacao.

TcPR-10, a member of the pathogenesis-related protein 10 family, was identified in EST library of interactions between Theobroma cacao and Moniliophthora perniciosa. TcPR-10 has been shown to have antifungal and ribonuclease activities in vitro. This study aimed to identify proteins that are differentially expressed in M. perniciosa in response to TcPR-10 through a proteomic analysis. The fungal hyphae were subjected to one of four treatments: control treatment or 30-, 60- or 120-min treatment with the TcPR-10 protein. Two-dimensional maps revealed 191 differentially expressed proteins, 55 of which were identified by mass spectrometry. The proteins identified in all treatments were divided into the following classes: cell metabolism, stress response, zinc binding, phosphorylation mechanism, transport, autophagy, DNA repair, and oxidoreductases. The predominant class was stress-response proteins (29%), such as heat shock proteins; these proteins exhibited the highest expression levels relative to the control treatment and are known to trigger defense mechanisms against cytotoxic drugs as well as TcPR-10. Oxidoreductases (25%) were overexpressed in the control and in 30-min treatments but exhibited reduced expression at 120 min. These proteins are involved in the repair of damage caused by oxidative stress due to the contact with TcPR- 10. Consistent with the antifungal activity of TcPR-10, several proteins identified were related to detoxification, autophagy or were involved in mechanisms for maintaining fungal homeostasis, such as ergosterol biosynthesis. These results show that the sensitivity of the fungus to TcPR-10 involves several biochemical routes, clarifying the possible modes of action of this antifungal protein.

Comparative study of putative 9-cis-epoxycarotenoid dioxygenase and abscisic acid accumulation in the responses of Sunki mandarin and Rangpur lime to water deficit.

Abscisic acid is a plant hormone that participates in essential plant physiological processes, especially during adaptation to many environmental stresses, such as water deficit. The relationship between ABA accumulation and the expression of putative carotenoid cleavage dioxygenase (CCD) genes was investigated in the pot-cultivated leaves and roots of the 'Rangpur' lime and 'Sunki Maravilha' mandarin plants. Transpiration, stomatal resistance and leaf growth were evaluated when these genotypes were subjected to continuous water deficit. Under water deficit conditions, the 'Rangpur' lime extracts used greater amounts of water when compared to the 'Sunki Maravilha' plants, which reached the greatest stomatal resistance 5 days before 'Rangpur' lime. When subjected to water deficit, the roots and leaves of 'Sunki Maravilha' showed a progressive increase in ABA accumulation; however, in 'Rangpur' lime, alternations between high and low ABA concentrations were observed. These results suggest a retroactive feeding regulation by ABA. In 'Rangpur' lime the NCED2, NCED3 and CCD4a genes were expressed at the highest levels in the roots, and NCED5 was highly expressed in the leaves; in 'Sunki Maravilha', the NCED2 and NCED5 genes were most highly expressed in the roots, and NCED2 was most highly expressed in the leaves. However, for both genotypes, the transcription of these genes only correlated with ABA accumulation during the most severe water deficit conditions. The 'Rangpur' lime behaved as a vigorous rootstock; the leaf growth remained unaltered even when water was scarce. However, 'Sunki Maravilha' adaptation was based on the equilibrium of the response between the root and the aerial tissues due to water restriction. The use of the Sunki mandarin in combination with a scion with similar characteristics as its own, which responds to water deficit stress by accumulating ABA in the leaves, may display good drought tolerance under field conditions.