Search

During the last 40 years a lot of mathematical models were developed for simulation of cardiomyocytes electrophysiology. In this work we propose an algorithm for meta-analysis of these models. Using pairwise comparison, our approach searches for outliers in model predictions that can be obtained in one model, but not in the other. Here, we analyzed four models of the atrial cardiomyocytes with the proposed approach. Our results show significant differences in the ability of models to simulate the delayed afterdepolarization, early afterdepolarization, sustained triggered activity, and the failure of activation. The proposed approach may be useful for the comparison of two models and for obtaining additional rejection criteria for studies with the model populations. Such criteria may improve the reliability of mathematical model predictions. [ABSTRACT FROM AUTHOR]

Pseudo electrocardiogram (pseudo-ECG) is a virtual ECG signal that is computed using 1D mathematical models of the myocardium. This signal has the QRS-complex and T wave similar to that of real ECG. In this paper, we used pseudo-ECG as a tool to develop an algorithm for the estimation of electrophysiological properties of the myocardium or single cardiomyocytes based on electrogram recordings. Here, we analyzed the relationship between pseudo-ECG produced by a 1D myocardial model and characteristics of the cellular action potential and parameters of an ionic cardiomyocyte model. We showed that in the noise-free conditions the action potential morphology and cardiomyocyte model parameters can be identified from pseudo-ECG using a linear transformation. [ABSTRACT FROM AUTHOR]

Brain tumor images segmentation plays a crucial role in the auxiliary diagnosis of disease, treatment planning and surgical navigation. In order to accurately segment brain tumor images, this paper proposes an automatic brain tumor Magnetic Resonance Imaging (MRI) image segmentation algorithm based on the U-net model. The neuron network based approached was closely analyzed in comparison with standard segmentation techniques (detection by threshold, K-means clustering, histogram-based approach and edge detection). The algorithm is validated and evaluated on the Jun Cheng dataset. The experimental results show that the proposed algorithm has strong competitiveness compared with the existing brain tumor MRI image segmentation algorithm. [ABSTRACT FROM AUTHOR]

The paper describes the application of the rapid serial visual presentation (SRVP) method to study the dynamics of recognition time of visual stimuli by an untrained observer. The method is implemented on the basis of a specially developed program presenting stimulus material on a computer screen according to a given algorithm. Raster monochrome images of two letters of the Russian alphabet, depicted in two significantly different styles, acted as stimulus material. Presentation was carried out in a specialized viewing camera. When evaluating the results of the experiment, ruggedness was taken into account as a quantitative measure of the lettering. In the experiment, an increase in the recognition time of all types of stimuli with an increase in the number of presentations was found. Using the program greatly facilitated the experiment. [ABSTRACT FROM AUTHOR]

An algorithm for coordinating commands for executing devices according to the "two out of three" reliability scheme, which can be used in technical systems, as a way to increase reliability when generating critical hardware level commands, is proposed. The developed algorithm has linear complexity. During the experiments, varying the number of executing devices from 1 to 100, an assessment was made of the number of coordination failures and the time required to coordinate one command, in the case when the communication line between logic controllers and executing devices is considered safe. [ABSTRACT FROM AUTHOR]

The convolutional neural networks with supervised training solve problems of semantic segmentation in a wide area of biomedical applications. However, this approach requires a significant size of the training dataset. Unsupervised learning can provide segmentation of some images without training dataset. Joint unsupervised learning is an approach that combines neural networks with another algorithm of data clusterization. In this preliminary study, we test the joint unsupervised learning approach with minor modifications for the problem of semantic segmentation at several datasets of biomedical images. [ABSTRACT FROM AUTHOR]

Using a reduced set of facial keypoints improves the accuracy of calculating the angles of rotation of the head. In this paper, we propose methods for evaluating the accuracy of calculating angles and an algorithm for selecting a subset of facial keypoints. This approach has improved the accuracy of angle calculation by more than 50%. [ABSTRACT FROM AUTHOR]

The problem of minimizing the time of reloading nuclear fuel for the BN-800 reactor is considered. Time- optimal control algorithm is proposed under the assumption that plugs can rotate sequentially. [ABSTRACT FROM AUTHOR]

Scientific idea of the modeling is based on assumption of similarity in crystallographic lattice transformations for shear phase and structural transformations. Occurrence of orientation relationship between parent and final phases is based on parallelism of high density packed planes and directions. This results in phase nucleation on suitable boundaries «substrate». In case of structure transformations, the special misorientations play the role of the orientation relationships, while special boundaries play the role of substrates. The result of the study is knowledge based algorithms of control over texture formation in metal based materials resulted from crystallographically oriented phase and structure transformations focused on achievement of required combination of operational properties in product. [ABSTRACT FROM AUTHOR]

Modern steel production facilities, such as rails, pipe billet (rod), and pipe, use various methods of non- destructive testing. The most promising method of non - destructive testing is magnetic flaw detection using anisotropic magneto resistive sensors of the AMRD type [1]. [ABSTRACT FROM AUTHOR]

In this paper, we propose an original approach to provide numerical stability and sustainability of stochastic schemes designed for solving multiparticle quantum spin models. It is presented as a combination of two essential improvements. The first one is refreshed view of stochastic energy estimation for an arbitrary quantum state, related to Heisenberg Hamiltonian. The second one appears as a modification of the genetic optimization algorithm to ensure stable convergence in actual computer calculation. The approach was tested on sample system, demonstrating efficiency and perspectives for further investigations in the field of microscopic magnetism. [ABSTRACT FROM AUTHOR]