Here we describe building a control impact intended to medically suppress tumor growth. The synthesis of control has been carried out using the method of linearization of a nonlinear system with state feedback. As a result of the study, a control law was obtained that provides the system with local stability, which translates into cessation of tumor growth in physical sense. The adequacy of the tumor growth model is achieved by constructing it using a self-organization algorithm with trend reservation. Linear trends are applied in the control law, while non-linear self-organizing models are used to test the treatment outcome. The results of mathematical modeling confirmed the effectiveness of the solutions obtained.

Causative relationships of microbiota with the human’s health and diseases are one of the most challenging issues in modern microbiology. Progress in this field could provide new tools for diagnosis, prophylaxis, and treatment. A new automated approach is proposed, as an addition to the methods of multidimensional statistical analysis. This approach is based on a planar projection of multidimensional analytical data and is distinguished by technological simplicity and clarity of the process of operational diagnostics.
Aim of the study was to apply a new automated approach based on the method of mapping diagnostic fields to determine the informative parameters and the main patterns of eubiosis / dysbiosis of the human large intestine, and the development of chronic prostatitis with fertility loss in men.
Materials and methods. Using the method of mapping diagnostic fields, whose geometrization is based on multidimensional observations of the state of each subject, the dimension of the feature space is determined by calculating the resultant of each feature vector and for calculating the Voronoi diagram - a diagnostic palette. Two samples were used in the work: the first consisted of 126 strains isolated from 65 individuals examined for colon dysbiosis (18–45 years old), the second consisted of 124 tests taken from 73 men of reproductive age (20–45 years old).
Results. The cartography method of the resultants creates easily interpretable graphic documents based on the initial data and contributes to the prompt recognition of unknown states/diagnoses. The created cartograms remove the limitations of perspective visualization of multidimensional objects and significantly simplify data interpretation.
Conclusions. The effectiveness of cartographic diagnostics has been confirmed by comparing its results with clinical ones. Both initial observations and statistically processed material can be used as data.

New measures to decrease the burden from cardiovascular morbidity are of great socio-economic importance. The aim of the study was to create artificial intelligence technology incorporating various methods and approaches for presenting and using knowledge to assess and predict individual risks of developing cardiovascular events. The following risk presenting models were used: scoring system, multivariate Weibull and logistic regression, artificial neural networks; an ontological approach for explicit representation of knowledge and the construction of software solvers generating an explanation in easy-to-interpret terms. One of the main technological solutions used was the IACPaaS cloud platform, which has infrastructure and intelligent service development technology. The result of the study is a hybrid technology for risk assessment and forecasting, presented in the article by the architecture of the decision support services produced, the ontology of knowledge, the knowledge base for cardiology and the methods for implementing services. The key feature of the technology is its scalability by connecting new microservices implemented on arbitrary heterogeneous architectures. The scope of application ranges from cardiology research of risk assessment and prognosis to medical practitioners.

Information technologies in modern continuous professional education (CPE) of medical workers are playing an increasingly active role, becoming the basis for the training organization. Aiming to substantiate the strategy for the further development of CPE, we conducted a questionnaire survey of doctors (n=211) immersed in new forms of training using distance learning technologies. Furthermore, regulatory documents were provided that should become a basis for creating conditions for effective training of doctors in real conditions of restrictions on full-time presence. Upon the results of the questionnaire, doctors’ beliefs and assessments were systematized. The most stated problem was the absence of the appropriate learning conditions in the workplace. To date, a legal basis was created for the training of doctors using distance learning technologies. The actual problem is how quickly both doctors and their employers will implement the new norms of the Labor Code in practice.

The problem of early and timely accurate diagnosis of rare hereditary diseases is global. The use of physician-assisted computer systems could solve it. There are various foreign medical decision support systems. However, there are currently no domestic functioning systems to resolve these issues.
The aim of this study is to improve the timeliness and accuracy of making a correct diagnosis in patients with signs of hereditary lysosomal storage diseases using an intelligent computer decision support system.
Materials and methods. To fill the knowledge base, various sources of medical information containing descriptions of the phenotypic manifestations of a group of lysosomal storage diseases were analyzed. The knowledge extracted from the literature was supplemented by three expert assessments — the coefficient of modality, the confidence measures of manifestation and degree of expression. For clinical approbation of the system, 35 clinical cases from the literature and depersonalized extracts from the electronic health records of 75 patients treated at four specialized medical organizations of the Russian Federation were used.
Results. The GenDiES expert decision support system for the differential diagnosis of orphan hereditary diseases has been developed. The knowledge base of the system is implemented on the IACPaaS cloud platform in the form of an ontological network. This made it possible to describe diseases considering expert assessments for four selected age periods and entering data from patients with suspected hereditary diseases. The comparative analysis algorithm was used to assess the similarity of the patient’s clinical features with expert descriptions. The accuracy of the test results was 88.18% for the differential diagnostic series of five hypotheses.
Conclusion. Implementation of the knowledge base in the form of an ontological model provided the GenDiES expert system with high efficiency at the stage of forming hypotheses at the pre-laboratory stage of diagnosis.

It is very important to balance the processes of creating the simplest and most effective predictive models in medicine. The predictors in the model determine its quality and practical relevance but selecting them is not always easy. The aim of the study is to compare different methods of prediction selection to create medical prognostic models.
Methods. We compare simple methods, such as correlation, predictor filtering based on basic statistics, and Hosmer-Lemeshow univariate analysis, with more complex methods often used in machine learning, such as recursive feature elimination, LASSO regression, and classification trees. The predictive models were built using the binary multiple logistic regression method. Statistical analysis was carried out using the programming language R (version 3.4.2).
Results. Based on the LASSO and random forest methods, as well as the stepwise regression method, the most accurate predictive models were constructed (minimum AIC value). The Hosmer-Lemeshow method and basic methods of statistical analysis have been found to be the least effective.
Conclusion. The use of predictor selection methods often significantly reduces their number, filtering out non-informative ones, which improves the quality of the predictive model.

There is a trend in the healthcare of the Russian Federation for the development and implementation of various information systems, including medical information systems of medical organizations (MIS MO), aimed automation of diagnostic, treatment, administrative, auxiliary and other processes, as well as radiological information systems (RIS), designed to automate the service of radiation diagnostics and radiology. Many of the currently utilized MIS were developed 10 and more years ago, are based on different (sometimes not compatible) technological and architectural approaches, have a number of problems with interface convenience, integration with existing healthcare information systems, speed and quality of support from the developer company. In this context medical organizations face the task of choosing MIS MO or RIS from the offers available on the market in order to replace the old system and / or implement a new solution. Without a unified methodology and evaluation criteria, such a choice is rather difficult and non-transparent. The authors of this article have developed and proposed a methodology that allows standardizing the assessment of the MIS MO and RIS maturity level to make decisions on the advisability of maintaining the current product or replacing it with solutions from other companies.