Departments
Home Departments
Directions of activity
Cooperation and services
- Development and development of models, methods of knowledge detection in molecular genetic, epidemiological, clinical, laboratory data of the medical and biological profile.
- Creation of special software of epidemiological and clinical registers of various diseases, information and analytical systems, decision support systems, implementing modern models developed in the laboratory, data analysis methods, machine learning algorithms for their use in healthcare.
- Support of special software of the integrated information system "Hybrid data bank of the Belarusian space system of remote probing of the Earth".
- Support of special software of the republican automated information system "Electronic Recipe".
- Support of special software of the information system "Register of children in need of palliative care".
- Accompaniment of special software of information and analytical system of post-transfusion reactions and complications.
- Official OOC within the framework of the creation of a centralized information system of health care of the Republic of Belarus.
Results of scientific research and development
- Development and validation of mathematical models of risk for the intellectual functional module for predicting the long-term consequences of treatment of patients of childhood with malignant neoplasms.
- Methods, biomedical data analysis models and information-analytical systems for prognosis of survival and diagnosis of socially significant diseases.
- The method of constructing a hybrid classification model, based on a combination of the bagging procedure and aggregation of ranked lists.
- The method of constructing an undirected gene regulatory network and the assessment of the statistical significance of intergenic connections using information criteria and technology of boottraping.
- Method of assessing the metastatic potential of malignant tumors and mathematical models of prognosis of the prevalence of the tumor process, allowing to predict the likelihood of detecting metastases of papillary thyroid cancer in the lateral group of lymph nodes for the spread of carcinoma in the lymph nodes of the central group, taking into account the metastatic index, which is considered as a risk factor.
- Method of integrating molecular genetic data to predict the functions of unknown proteins.
- Algorithm of fuzzy clustering of genetic data with partial control and active selection of constraints, within which analytical expressions are obtained for the parameters of the extended optimization function, a procedure for active selection of restrictions has been developed, based on the definition of the “virtual” boundary of clusters and the choice of restrictions for pairs of points that fall into this boundary, and comparative tests have shown that the algorithm allows to improve the quality of data clustering with fewer restrictions.
- Mathematical models for detecting the morphological features of papillary thyroid cancer, which determine the invasive and metastatic potential of the tumor in children and adolescents in the Republic of Belarus, developed on the basis of a retrospective clinical and morphological analysis of papillary thyroid cancer in children and adolescents, whose occurrence is associated with an “iodine impact”, received by patients in the first months after the accident.
- The tracing rank method (biomarkers) by reassessing the effectiveness of the prediction of the phenotype of the disease for randomly generated samples from gene expression data.
- Nonparametric algorithm for clustering gene expression profiles based on an automatic assessment of the local density function with the representation of the result in the form of a hierarchical cluster structure.
- Nonparametric method of clustering genetic data, based on the concept of stability, allowing to determine the number of clusters in data and the stability of individual clusters and assess the statistical significance of the resulting clustering results.