Biotekhnosfera №2/2018

Khachatryan K. S., Manilo L. A. Methods for assessing the informativeness of signs in the formation of a morphological description of a pseudo-phase portrait of rhythmograms

The article considers several approaches to assessing the informative value of the morphological

features of a pseudo-phase portrait in the problem of arrhythmia recognition by rhythmogram. The possibilities of reducing the dimensionality of the feature space on the basis of the information approach and the application of the cluster analysis procedures are shown. In the course of experiments performed on real rhythmogram records, it was established that the construction of silhouette graphs and analysis of the numerical value of the silhouette allows objectively to evaluate the effectiveness of applying the obtained set of characteristics for solving a multiclass

problem.

Kalinichenko A. N., Kalinin N. G., Al-Ghily M. A. Anesthesia depth estimation using electroencephalogram on the base of linear discriminant methods

The main goal of the presented investigation was development of an algorithm for the anesthesia

depth estimation using electroencephalogram (EEG) that can be used in the monitors for the anesthesia stages control in the course of surgical operations. A set of nine indices calculated from EEG signal is used as input data for the linear discriminant analysis procedure. It was shown that combination of all nine analyzed parameters provides best differentiation between various stages of anesthesia. All parameters were ranked also according to their impact to the total algorithm performance.

Sadykova E. V. Development of a web site for diagnosis of rare diseases with decision-making under uncertainty

Rare occurrence of orphan diseases carries little information content, which leads to difficulties in obtaining timely and correct diagnosis. Therefore, the remote diagnostic system for collecting data on rare diseases is required. Work contains: a formulation of the problem, a review of methods for formalization and medical data processing, prototyping and designing the structure of a web site and program development. As a result of execution of the work we developed a web site for diagnosis of rare diseases with decision support system for doctor and information support for patient. The developed system can be widely used for diagnostics of rare diseases.

Dolganov A. Yu., Kublanov V. S. Biotechnical system for treatment of diseases accompanying disorders of the autonomic nervous system regulation

The article describes key elements of the biotechnical system for treatment of diseases accompanying disorders of the autonomic nervous system regulation: informational decision support system and neuro-electrostimulator. As a clinical model arterial hypertension was selected. Informational decision support system is based on the application of the machine learning methods to the diagnostically significant sets of heart rate variability features. Structure of neuro-electrostimulation services which allows interaction of the neuro-electrostimulator and informational decision support system is proposed.

Smirnova L. M., Monakhova M. I., Yuldashev Z. M. Assistive technology for personified synthesis of modular lower limb prostheses

The article discusses the problematic issues of personified synthesis of lower limb modular prostheses from unified functional modules (components). Attention is paid to the need to develop a technology of such synthesis based on the analysis of the functioning of the prosthesis in the bioengineering system “patient – prosthesis – environment”. The stages of technology development and algorithms of personalized synthesis of modular prosthesis in the work of prosthetic and orthopedic enterprise are indicated.

Grecheneva A. V., Dorofeev N. V., Kuzichkin O. R.

Joint proceessing of goniometric and electrophysiological signals in the automated system of diagnostics and rehabilitation of the musculoskeletal sysem

The article suggests an approach of neural network processing of accelerometric and electrophysiological signals and construction of an adaptive individual patient model. Correlation links in the system “nervous system — musculoskeletal system” have been revealed and described. A generalized model of adaptive data processing is presented, which includes models of subsystems for collection and processing, initialization and adaptation, estimation, prediction, harmonization and control. A generalized structure of the neural network for processing the recorded data is presented. Results of modeling and training of the obtained neural prediction network for detecting tremor of the brush are presented.

Yuldashev Z. M., Nguyen Ch. T., Chan Ch. H., Nguyen M. T. An algorithm of intelligent biomedical signals processing and analysis for the human health status remote monitoring system

In the health status remote monitoring systems of people with chronic diseases in order to avoid skipping episodes of exacerbation of the disease, longterm continuous work is of great importance. The autonomy of the work of such systems is largely determined by the autonomy of the patient’s wearable elements of the system. To ensure long-term continuous work and efficient functioning of the monitoring system, it must have a hierarchical structure with elements realizing recording and picking off biomedical signals, signal processing and analysis, estimation of the current health status, dynamics of the disease and prognostication of the human health status. To ensure long-term work of the system, it is necessary to use intelligent monitoring algorithms. The peculiarity

of such algorithms is the change in the number of channels for recording and biomedical signals processing depending on the change in the human health state. In detecting the first signs of exacerbation of the disease with the patient’s wearable computer, additional channels for recording biomedical signals used to evaluate the expanded complex of diagnostically significant parameters of the disease and their integration during the classification of the patient’s health state are activated. An example of the system construction and the development of an intelligent monitoring algorithm for the atrial fibrillation episodes detection system is investigated.

Korolev D. V., Shulmeister G. A., Romanova T. N., Postnov V. N. Investigation of the effectiveness of various organic solvents in the synthesis of an aminespacer on magnetic nanoparticles

The analysis of various solvents used for amination of the surface of magnetic nanoparticles based

on magnetite (Fe3O4) using (3-aminopropyl) triethoxysilane (APTES) was carried out. As solvents used: benzene, toluene, cyclohexane, isopropanol, ethanol. The total content of amino groups in the samples was determined by inverse titration, the number of available amino groups by the capacity of indocyanine green (ICG). A comparison of the results obtained showed that the best combination of the concentration of amino groups and ecology is observed in the synthesis of cyclohexane, and the lowest concentration of amino groups is in the synthesis from alcohols, and in this group from isopropanol.