Iveta Mrázová

Knockout of Angiotensin 1–7 Receptor Mas Worsens the Course of Two-Kidney, One-Clip Goldblatt Hypertension: Roles of Nitric Oxide Deficiency and Enhanced Vascular Responsiveness to Angiotensin II

Aims: The present study was performed to evaluate the effects of target disruption of the G-protein-coupled receptor Mas for angiotensin 1–7 [Ang(1–7)] in knockout mice on the course of two-kidney, one-clip (2K1C) Goldblatt hypertension. Methods: Knockout and wild-type mice underwent clipping of one renal artery. Blood pressure (BP) was monitored by radiotelemetry. The mice were either untreated or chronically treated with the superoxide (O2–) scavenger tempol (400 mg/l) or the inhibitor of NADPH oxidase apocynin (1 g/l) administered in drinking water. Results: Knockout mice responded to clipping by accelerated increases in BP and the final BP was significantly higher than that in wild-type mice. Chronic treatment with tempol or apocynin elicited similar antihypertensive effects in 2K1C/knockout as in 2K1C/wild-type mice. Acute nitric oxide synthase inhibition caused greater BP increases in 2K1C/wild-type than in 2K1C/knockout mice. Conclusion: Our present findings support the notion that the angiotensin-converting enzyme 2-Ang(1–7)-Mas axis serves as an important endogenous physiological counterbalancing mechanism that partially attenuates the hypertensinogenic actions of the activated renin-angiotensin system. The impairment in this axis may contribute to the deterioration of the course of 2K1C Goldblatt hypertension.

Persistent antihypertensive effect of aliskiren is accompanied by reduced proteinuria and normalization of glomerular area in Ren-2 transgenic rats

The effects of the human renin inhibitor aliskiren on blood pressure (BP), end-organ damage, proteinuria, and tissue and plasma angiotensin (ANG) II levels in young and adult heterozygous Ren-2 transgenic rats (TGR) were evaluated and compared with the effect of the ANG type 1 (AT(1)) receptor blocker losartan during treatment and after 12 days after the withdrawal of drug treatments. BP was monitored by telemetry from the age of 32 days on (young rats) and at 100 days (adult rats). Aliskiren (10 mg·kg(-1)·day(-1) in osmotic minipumps) or losartan (5 mg·kg(-1)·day(-1) in drinking water) treatment was applied for 28 days in young rats and for 70 days in adult rats. In young untreated TGR, severe hypertension rapidly evolved. Adult untreated TGR exhibited stable established hypertension. Both aliskiren and losartan fully prevented the development of hypertension and cardiac hypertrophy in young TGR and normalized BP and cardiac hypertrophy in adult TGR. After cessation of aliskiren treatment in both young and adult TGR BP and cardiac hypertrophy were persistently reduced, while after losartan withdrawal BP and cardiac hypertrophy rapidly increased. In adult aliskiren-treated rats proteinuria was significantly reduced compared with losartan (the effect persisting after withdrawal of treatment), and this decrease strongly correlated with normalization of glomerular size in these animals. In conclusion, aliskiren and losartan had similar antihypertensive effects during chronic treatment, but the antihypertensive and organoprotective effects of aliskiren were persistent even after the 12-day washout period. The durable effect on proteinuria can possibly be attributed to the normalization of glomerular morphology.

Improved generalization of neural classifiers with enforced internal representation

In standard BP-networks, hidden neuron outputs are usually spread over the whole interval (0,1). In this paper, we propose an efficient framework to enforce a transparent internal knowledge representation in BP-networks during training. We want the formed internal representations to differ as much as possible for different outputs. At the same time, the hidden neuron outputs will be forced to group around three possible values, namely 1, 0 and 0.5. We will call such an internal representation unambiguous and condensed. The performance of BP-networks with enforced internal representations will be examined in a case study devoted to semantic image classification.

Can Deep Neural Networks Discover Meaningful Pattern Features

Abstract Recent advances in the area of deep neural networks brought a lot of attention to some of the key issues important for their design. In particular for 2D-shapes, their accuracy has been shown to outperform all other classifiers - e.g., in the German Traffic Sign competition run by IJCNN 2011. On the other hand, their training may be quite cumber- some and the structure of the network has to be chosen beforehand. This paper introduces a new sensitivity-based approach capable of picking the right image features from a pre-trained SOM-like feature detector. Experimental results obtained so far for hand-written digit recognition show that pruned network architectures impact a transparent representation of the features actually present in the data while improving network robustness.

Hybrid convolutional neural networks

Convolutional neural networks are known to outperform all other neural network models when classifying a wide variety of 2D-shapes. This type of networks supports a massively parallel extraction of low-level features in the processed images. Especially this characteristic is assumed to impact the performance of convolutional networks in character recognition tasks - and in particular when considering scaled, rotated, translated or otherwise deformed patterns. Yet training of convolutional networks is rather time-consuming due to the relatively high complexity of the entire model. To speed-up the training process, we will propose a new variant of convolutional networks - the so-called hybrid convolutional neural network (HCNN). HCNN-networks combine the original idea of LeCuns convolutional networks with the benefits of RBF-like neurons in all the layers and with the winner-takes- all mechanism applied during recall. In the tests done so far in hand-written digit recognition, HCNN proved to be capable of considerably speeding-up the training process while maintaining roughly the same performance of the trained networks like original convolutional networks.

A new sensitivity-based pruning technique for feed-forward neural networks that improves generalization

Multi-layer neural networks of the back-propagation type (MLP-networks) became a well-established tool used in various application areas. Reliable solutions require, however, also sufficient generalization capabilities of the formed networks and an easy interpretation of their function. These characteristics are strongly related to less sensitive networks with an optimized network structure. In this paper, we will introduce a new pruning technique called SCGSIR that is inspired by the fast method of scaled conjugate gradients (SCG) and sensitivity analysis. Network sensitivity inhibited during training impacts efficient optimization of network structure. Experiments performed so far yield promising results outperforming the reference techniques when considering both their ability to find networks with optimum architecture and improved generalization.

Semantic Clustering of the World Bank Data

World Development Indicators (WDI) published annually by the World Bank provide comparative socio-economic data for state economies. Several countries show common trends in their development. But to understand these trends in the development process, an appropriate interpretation of the intrinsic similarities has to be found. In this paper, we propose a novel approach to assigning an adequate semantics to clusters formed by fuzzy c-means clustering. Despite of the ability to identify unique characteristics for the found clusters, the introduced fuzzy c-landmarks show a great potential for dimension reduction and for simplified data set descriptions. Experiments performed so far confirm efficient processing for this kind of exploratory data analysis.

Czech Insolvency Proceedings Data: Social Network Analysis☆

Abstract In 2008, the Czech government launched an information system called Insolvency Register of the Czech Republic. Already within the first year of its operation, about 5200 insolvencies were commenced. This number rose, however, immensely in the following years as the impact of the Global Financial Crisis became evident also in the Czech Republic - industrial production fell by 13.4% and many areas witnessed massive layoffs. Meanwhile, the Czech Insolvency Register contains publicly available data concerning approximately 160 000 insolvency proceedings. The subjects participating in insolvency proceedings include debtors, creditors, senates deciding in the respective insolvency matter and insolvency administrators who handle debtors assets during the proceedings. Often, the involved subjects participate in several insolvency proceedings thus forming a complex social network that evolves over time. In order to better understand emerging trends in such a type of networks, we orient our research towards the identification of influential individuals. Association rules used to predict future behavior of the network reveal several interesting patterns present across the entire country as well as locally in specific regions only.

Fast and Reliable Detection of Hockey Players

Abstract Current popularity of augmented reality (AR) stems from its ability to enhance the perceived environment in real- time with additional information of semantic context, such as sports scores shown on TV during match broadcasting. Its other application areas range from industry and medicine to military, commerce and entertainment. Advanced AR technologies obviously rely on accurate, yet fast enough algorithms for multimedia processing and object recognition. In this paper, we will study the possibility of using convolutional neural networks (CNNs) for real-time detection of hockey players from video streams of broadcasted ice-hockey matches. Supporting experiments performed so far yield sufficient accuracy for this task (above 98.5%), while maintaining reasonable computational demands and acceptable robustness both with regard to noise and minor image transformations like translation, rotation and scaling.

Sensitivity-based SCG-training of BP-networks*

Abstract Reliable neural networks applicable in practice require adequate generalization capabilities accompanied with a low sensitivity to noise in the processed data and a transparent network structure. In this paper, we will introduce a general framework for sensitivity control in neural networks of the back-propagation type (BP-networks) with an arbitrary number of hidden layers. Experiments performed so far confirm that sensitivity inhibition with an enforced internal representation significantly improves generalization. A transparent network structure formed during training supports an easy architecture optimization, too.