Intelligent Analysis of Medical and Psychophysiological Data (invited paper) Nafisa Yusupova and Konstantin Mironov Faculty of Computer Science and Robotics Ufa State Aviation Technical University, Ufa, Russia yussupova@ugatu.ac.ru, mironovconst@gmail.com Abstract The paper is dedicated to the application of Intelligent methods of data analysis on the examples of medical and psychophysiological tasks. Al- though there are pretty much research in this field, unified complex methodology of medical data analysis does not exist. In this paper we present the short overview of using various means of data analy- sis in medical applications: big data, machine learning, text mining, multi-agent systems. We present wo cases of intelligent data analysis performed by the researchers from Ufa State Aviation Technical Uni- versity in collaboration with experts and researchers from the medical institutions in the city of Ufa. First case consist in analysis of weak- structured data about acute poisonings in the Republic of Bashkor- tostan. The second case was connected to analysis of the results of psychophysiological diagnostics of students in order to determine rec- ommendations for their physical activity. 1 Introduction In recent years several regulations was made in the sphere of medical data processing, e.g. [Ord11] and [Dec13]. According to these documents, a unified information system was introduced in medical organizations of Bashko- rtostan Republic. More than five years have passed since then, and a lot of medical data has been accumulated, in particular 4 million electronic patient records; 346 million cases of medical care; 370 thousand bills for the medical care; over 35 million medical images; 13 million referrals for laboratory research; 9 million electronic recipes. The accumulated dataset is a typical object of big data technology, containing a combination of un- structured and poorly structured knowledge about the processes and methods of treatment. This point of view allows us to consider the accumulated mass of information as a valuable resource for analytical work. In this regard, it is of interest to analyze related works in the field of medical big data. Although there are pretty much research in this field, unified complex methodology of medical data analysis does not exist. In this paper we present a short overview of existing related works on medical data analysis (section 2), and an two cases of medical data analysis performed by researchers at the Faculty of Computer Science and Robotics, Ufa State Aviation Technical University. Copyright 2019 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). In: S. Hölldobler, A. Malikov (eds.): Proceedings of the YSIP-3 Workshop, Stavropol and Arkhyz, Russian Federation, 17-09-2019–20-09-2019, published at http://ceur-ws.org 1 2 State of the Art There are a lot of works in the field of big data in recent years. However, there are not so many publications in the open press on the application of this approach to the study of medical data. In this area, the following significant studies can be noted [Cve16,Il16,Eni18,TOM,Tha19,Ron14,Jak16]. In the works [Cve16,Il16], it is proposed to distinguish the following groups of practically significant tasks: quick identification of patients with various risks; increasing the effectiveness of medical interventions; making the best decisions; close monitoring; comparison of significant clinical data with the results of Big Data. Moreover, up to 90% of the data is unstructured. As a rule, data in different institutions are presented in various formats. Information comes from various sources and from various clinical systems. A 44-fold increase in data over the next decade is expected (one exabyte by 2020). The use of data mining for the analysis of medical data has been the subject of many works, in particular, publications [Ugl17,Aks18,Sai18,Lin16,Muh15]. In these works, the use of the following analysis tools was studied: classification, modeling and forecasting methods based on the use of decision trees, artificial neural networks, genetic algorithms, evolutionary programming, associative memory, fuzzy logic. In [Sai18], an analysis of the used methods and technologies of data mining in the field of healthcare is presented. The authors of [Lin16] propose using data mining methods to solve public health management problems. The study [Muh15] is devoted to the tasks of data mining from a smartphone and wearable devices. An analysis of the studies showed that machine learning methods are usually used to improve the analysis of visual data and images, for example, in the works [Sha17,Zho17]. Paper [Cir12] is devoted to the application of deep learning for solving the task of classifying medical images. The following basic technologies are used in the considered papers: NoSQL (DBMS with non-relational data structure); Hadoop (the technological core of the project ecosystem for working with data); MapReduce (distributed computing model for parallel processing of large amounts of data; implemented in the Hadoop system); R language (programming language for statistical data processing and graphics); Python (a language for scientific computations with fine ecosystem of libraries, modules and applications). The task of automatic analysis of medical unstructured texts is relatively new and relevant [Gal17,Men15,Tch10,Ros10], but there are no ready-made solutions in this area. Questions of automatic text analysis can be related to the construction of ontologies. There are tools for automatic ontol- ogy generation based on structured [Kur17] and unstructured material [Orb12,Moz11,Mas14,Kum16,Piv05]. The general questions of knowledge formalization are the subject of many works, in particular, [Gav03,Vas17,Nov18,Rai18,Sam09,Tra97,Yat82,Pop96,Mur07,Lop04]. However, the specificity of subject areas requires additional research and formalization of knowledge. There are some works on the formalization of knowl- edge in the field of medicine [Aba13,Kot05]. The authors of [Ber12] show that it is possible to present operational definitions of diseases using OWL and to successfully classify real cases of patients. A feature of the ontology de- veloped in [Kha09] is the inclusion of temporality. The authors of [Ald17] suggest the need for additional research to identify bad practices and anomalies in the development of ontologies by computer scientists by the medical profession. Patients’ data used belong to the category when it is necessary to save both the past and current state of the database, therefore, it is necessary to consider data temporality [Kos07,And98,Eli12,Kol09,Baz09,Koz10]. The application of the multiagent approach in healthcare was considered, for example, in [Wit04]. The developed multi-agent system simulates the interaction of general practitioners, the chief physician of the clinic, specialists of the hospital, ambulance, medical university, managers of the Ministry of Health of the region, the territorial fund of compulsory medical insurance, an authorized pharmacological enterprise, and a patient of medical institutions. In [Dor15], a review of the use of multi-agent systems for various health problems is given. The authors identify the following areas of application of multi-agent systems: study of the effectiveness of different mechanisms of interaction of agents, different scheduling heuristics; operational planning of the treatment process; building simulation models of a specific hospital. Such a model reproduces with maximum accuracy the organizational structure of the hospital (or its parts), resources (wards, beds, equipment, staff), the interaction mechanism of the units and the real statistical characteristics of the patient flow. This model allows you to improve the organization of the healing process. In addition, the following areas of application of multi- agent systems in healthcare can be distinguished: decision support for managers, balancing and maximizing the use of available resources [Ben15]; DSS for managing hospital resources [Nes18]. The analysis of the related works has shown that work in the field of application of intelligent technologies for processing medical data is actively carried out in different countries and in different directions. However, the methodological basis for the formation of intelligent decision making for diagnosis, treatment and further support of the patient, combining a variety of intelligent technologies in a single methodology, is not sufficiently developed, therefore this problem is fundamental, and its solution is relevant and practically significant. 2 3 Research in Ufa The scientists from the Faculty of Computer Science and Robotics at Ufa State Aviation Technical University made significant fundamental and practical research work in the field of medical data analysis in collabora- tion with experts from Bashkir State Medical University, Bashkortostan Kuvatov Republican Clinical Hospital number 21, Ufa City Clinical Hospital, and the chair of Physical Education at Ufa State Aviation technical University. In this article we describe two example cases of data analysis: exploration of the acute poisoning in the Republic Bashkortostan and decision support for improving the psychophysical readiness of students for successful professional activities. Case 1: analysis of toxicologic data in Bashkortostan Republic A group of scientists from Ufa State Aviation Technical University (lead by professors Nafisa Yusupova and Gouzel Shakhmametova) together with coleagues from Bashkortostan State Medical University (lead by professor Rustem Zulkarneev) has made a research [Yus18] on the toxicological data from the Republic of Bashkortostan for 2015-2016. The goal was to construct and apply a complex technique for the analysis of toxicological data including methods of mathematical statistics and data mining. The analysis of the data about the cases of poisoning could support decision making for treatment and prevention of toxicological diseases. These decisions are important not only from the medical, but also from the social point of view. This allow one to carry out the comprehensive analysis and to benefit from the largest possible amount of knowledge, interrelations and patterns. The input of the data processing module included 6338 diversed records about the cases of poisoning in unstructured and semistructured form. The requested output include the following information: main reasons and structure of acute poisonings, structure of poisons, dependance on age and gender, definition of poisoning outcomes, etc. Use of data mining allowed discovering patterns among large volumes of data, which are objective and practically useful but invisible for statistical analysis. Parametric and non-parametric methods of statistical analysis were applied for processing quantitative data. Main results are presented in [Yus18] including some unexpected outcome about the structure of poisonings. E.g. main reasons of acute poisonings are the following: Alcohol (47,80%); Drugs (37,88%); Narcotic substances (5,99%); Carbon monoxide (5,43%); Mushrooms (2,15%); Snake bites (0,74%). Structure of the poisons which have caused acute poisonings is the following: Alcohol (28,9%); Carbon monoxide (49,2%); Narcotic substances (7,2%); Corroding substances (1,6%); Organic solvents and aromatic hydrocarbons (0,2%); Drugs (1,8%); Pesticides(0,001%); Other unspecified substances(11,1%). Exploration of poisoning dependence on age and gender showed the following results. For the children (age 015) no specific difference in poisoning was found. Adult men are poisoned more often than adult women: for the age 1630 67% of poisoned people are men. For the age 3145 this rate is 72%, for the age 4660 75%, and for the age 6175 70%. 60% of poisoned old people (age more the 75) are women; this may be explained by the fact that for this age total number of women is much more than total number of men. This case show that intelligent data analysis provide results, which are interesting from the theoretical point of view and may support decision making in the health-care management institutions. Case 2: Data Mining to support decisions on improving the psychophysical readiness of students for successful professional activities Researchers from Ufa State Aviation Technical University (group from the department of Computational Math- ematics and Cybernetics lead by professor Nafisa Yusupova and professor Olga Smetanina and expert Tatyana Naumova from the department of Sports Education) has explored the data about psychophysical conditions of students from the Faculty of Computer Science and Robotics [Yus19]. Human-machine interoperability plays an important role in the Industry 4.0 concept. To implement this concept, employees, including particular program- mers, will require psychophysical readiness. Purposeful psychophysical training of specialists is possible using a special model (professiogram), which includes a detailed description of the conditions and specifics of work [Ego05a]. Professions with increased requirements for psychophysical readiness require a mathematical model that takes into account the relationship of qualifications, professionally important qualities and their mutual influence [Ego05b]. The author of [Ego05b] also notes that mathematical modeling of higher mental functions allows one to purposefully choose means of physical education and sports in order to form the psychological readiness of future specialists for extreme working conditions. Sharopin [Sha07] has developed an information system for assessing students’ psychophysical readiness for professional activity, which allows one to obtain an integrated assessment of professional applied physical readiness. Sharopin also indicated that a quantitative 3 determination of the level of psychophysical readiness is necessary [Sha11]. Pichurin [Pic14] describes the role of physical education in the development of psychological and psychophysical preparation of students for profes- sional work. The analysis of the related works allowed us to conclude that it is possible to apply data mining and use the results to support decisions in this area. Methods for assessing professionally important physical qualities and mental properties are considered in [Sme16, Sme18]. Special tests, such as the Schulte test and the Rissou test, allow one to evaluate professional characteristics. For the development and improvement of professionally important physical qualities and mental properties, there is a certain composition of exercises. E.g., the following groups of physical exercises help to develop coordination abilities: exercises on the coordination of movements; exercises on the accuracy of movements; exercises in jumps and turns. Coordination exercises contribute to the development and improvement of psychological qualities such as attention, thinking and memory, so they must be developed together. As a rule, recommendations are given to a certain group of students with close values of indicators. Groups are defined by clustering. The formal statement of the task is as follows: it is necessary to identify groups of students with close values of indicators (test results) in order to develop general recommendations for improving psychophysical properties. Our methodology includes four steps. The first step is aimed at preparing data for analysis. The preparation tools are data cleaning algorithms (detection of anomalies, filling in gaps, identifying duplicates and contradictions). At the second step, clustering by Kohonen neural network is applied to identify the similarity of objects. At the third stage, recommendations are made in the form of a set of exercises for each cluster. In addition to the results of clustering, it was proposed to use the knowledge of experts. At the final step, the formed production knowledge base is used. To implement the methodology, a comprehensive analytical platform Deductor Studio was used. The results are given in table 1. This results allow one to define recommendations for students. Table 1: Defined clusters from [Yus19] Cluster Tapping test ”Hook” test Schulte test ”Running Yarotsky test (characteristic (static strength (characteriza- to numbered (general equilib- of the nervous endurance of the tion of volume, places” test rium)” system) hand muscles) distribution and (spatial ori- switching of entation and attention) memory) 0 160..180 40..180 28..40 9,75..10,7 19..39 1 165..190 30..63 22..30 7,8..8,8 30..43 2 155..180 105..145 22..40 7..8,8 39..50 3 150..170 45..105 34..47 7,9..9,7 23..39 An analysis of the results shows that students who are in cluster 2 can perform the basic set of exercises. Students who make up cluster 1 should perform exercises on the static strength endurance of the muscles of the hands. For a small part (those with a Yarotsky test of less than 39), balance exercises are also added. Students included in cluster 0 are characterized by a complex similar to the previous cluster. In addition to this complex, exercises on spatial orientation and memory are necessary. Cluster 3 turned out to be the most difficult group. In this case, it is necessary to compose a complex of exercises that contributes to the improvement of all characteristics. 4 Conclusion Research work in the field of intelligent technologies is being actively developed in such directions as big data, machine learning, text mining, multi-agent systems, etc. Different cases of applying these methods for the tasks of healthcare push their development and provide relevant results, which improve effectiveness of decision making in the field of healthcare. The researchers from Ufa State Aviation Technical University have long experience of scientific work in cooperation with researchers and practitioners from the field of healthcare. Achieved results can improve work conditions for medical workers and provide the institutions of healthcare with useful information. Considered cases of collaborative research show that medical data analysis include both classical (parametric and non-parametric statistical analysis) and intelligent methods for data clustering, classification, etc.. The second considered case demonstrate that means of intelligent data analysis allow one to define practical recommendations 4 on sports activity based on individual psychophysiological properties. Aquired experience is interesting from the scientific and practical points of view. It is also a promising starting point for further works. Acknowledgements The reported study was funded by RFBR according to the research project 18-07-00193. References [Aba13] Abayan A.V. Methods of formalizing and presenting knowledge about the subject area and the processes taking place in it when developing information systems for monitoring and analysis of complexly struc- tured dynamic medical data. Electronic Scientific and Educational Bulletin. Health and education in the 21st century. 2013.V. 15 (12) (in Russian). [Aks18] S. V.Aksenov, K. A.Kostin, A. V.Ivanova, J. Liang, A. V.Zamyatin Diagnosis of pathologies according to video endoscopy using an ensemble of convolutional neural networks. STM 2018 - Volume 10, No. 2 (in Russian). [Ald17] Aldosari, Bakheet&Alanazi, Abdullah&Househ, Mowafa DPitfalls of Ontology in Medicine. Studies in health technology and informatics. 238. 2017. pp. 15-18 [And98] Andreas Steiner A Generalisation Approach to Temporal Data Modelsand their Implementations. 1998. http://www.timeconsult.com/Publications/diss.pdf/ [Baz09] Bazarkin A.N. Development of a temporal data model in a medical information system. Software Products and Systems, 2009, No. 2, p. 3440. (in Russian). [Ben15] N.Benhajji, D.Roy, D.Anciaux Patient-centered multi agent system for health care. IFAC- PapersOnLine, Volume 48, Issue 3, 2015, Pages 710-714. [Ber12] Bertaud, Valerie & Duvauferrier, Regis & Burgun, Anita Ontology and medical diagnosis. Informatics for health & social care. 2012. No37. p.51-61. 10.3109/17538157.2011.590258 [Cir12] Ciresan Dan, Meier U., Schmidhuber J. Multi-column deep neural networks for image classifica- tion. June 2012 IEEE Conference on Computer Vision and Pattern Recognition, pp. 3642 to 3649. DOI:10.1109/cvpr.2012.6248110. [Cve16] L. A.Cvetkova, O. V.Cherchenko Implementation of Big Data Technologies in Healthcare: Assessing Technological and Commercial Prospects. Economics of science, 2016, vol.2, pp. 138 to 150 (in Russian). [Dec16] Decree of the Government of the Republic of Bashkortostan No. 376 dated August 14, 2013 On the Republican Medical Information-Analytical System of the Republic of Bashkortostan (http://docs.cntd.ru/document/463507025) (in Russian). [Dor15] Dorofeyuk A.A., Dorofeyuk Yu.A., Mandel A.S., Pokrovskaya I.V., Chernyavsky A.L. Multi-agent systems in the management of the treatment process in large-scale medical complexes. VIII International Conference ”Management of the development of large-scale systems” MLSD2015, Moscow, IPU RAS, 2015 (in Russian). [Ego05a] Egorychev E.A. Choice of criteria for managing students psychophysical readiness for professional activities. Physical Culture. Scientific and methodological journal. 2005. http://sportlib.info/Press/FKVOT/2005N1/p51-55.html (in Russian). [Ego05b] Egorychev E.A. Theory and technology of controlling the psychophysical preparation of students for professional activities. Theses for the degree of Doctor of Pedagogical Sciences, I.M. Russian State University of Oil and Gas Gubkina, Yaroslavl, 2005 (in Russian). [Eli12] Eliseev D.V., Baldin A.V. A review of methods for constructing temporal systems based on a relational database. Engineering Journal: Science and Innovation. 2012. No. 3. (3). pp. 5 - 12. (in Russian). 5 [Eni18] deputy Minister of Health of the Republic of Bashkortostan D. Enikeeva Practice of Health Informati- zation in Bashkortostan. Conference Practice of Health Informatization on the operation and develop- ment of regional segments of the Unified State Health Information System, built on the basis of RIAMS ProMed, April 5-6, 2018, https://health.bashkortostan.ru/presscenter/news/25886/ (in Russian). [Gal17] Galkina S. F. Institutional medical text: the experience of linguistic analysis. Philological sciences. Questions of theory and practice. 2017. No. 12 (78). Part 2. pp. 86-90 (in Russian). [Gav03] Gavrilova T. A. Ontological approach to knowledge management in the development of corporate infor- mation systems. J. ”News of Artificial Intelligence”, No. 2, 2003. p.24-30 [Gri18] V. V. Gribova, M. V. Petryaeva, D. B. Perch, E. A. Shalfeeva Ontology of medical diagnostics for intelligent decision support systems. Design Ontology, vol. 8, No. 1 (27), 2018. P. 58-73 (in Russian). [Il16] N. Yu.Ilyasova, A. V.Kupriyanov, R. A.Paringer FEATURES OF USING BIG DATA TECHNOLO- GIES IN TASKS OF MEDICAL DIAGNOSTICS. INTELLIGENT INFORMATION PROCESSING IIP-2016, abstracts of the 11th international conference, 10-14 october 2016, pp. 198 to 199 (in Russian). [Jak16] Jake Luo, Min Wu, Deepika Gopukumar, Yiqing Zhao Big Data Application in Biomedical Research and Health Care: A Literature Review. Biomed Inform Insights. 2016; 8: 110. Published online 2016 Jan 19. doi: 10.4137/BII.S31559 [Kha09] Khashaev Z.H. - M., Plesnevich G.S., Sheksheev E.M. Ontologies of medical knowledge with temporal aspects. Fundamental research. 2009. No. 2. pp. 51-55. (in Russian) [Kol09] Kolykhalova E.V., Proskurin D.K. Methods of designing information systems taking into ac- count the temporality of data in the subject area. III International Scientific Conference ”Mod- ern Problems of Informatization in Modeling, Programming and Telecommunications”, 2009. URL: http://econf.rae.ru/article/4796/ (in Russian) [Kos07] Kostenko B.B., Kuznetsov S.D. History and current problems of temporal databases. Electronic resource, 2007. Access mode: http://citforum.ru/database/articles/temporal/4.shtml (in Russian) [Kot05] Kotov Yu.B. Methods of formalizing the professional knowledge of a doctor in the problems of medical diagnosis. Doctor and information technology. 2005. No1. p. 62-68 (in Russian) [Koz10] Yu.V. Kozada, A.N. Bazarkin Data temporality issues in the integration of medical information systems. Program systems: theory and applications: electron. scientific journal 2010. No. 4 (4), p. 45-52. URL: http://psta.psiras.ru/read/psta2010 4 45-52.pdf (in Russian) [Kum16] N. Kumar, M. Kumar, M. Singh Automatic generation of ontologies from plain text using statistical and NLP methods. International Journal of Systems Support for Design and Management, 2016, Volume 7, Supplement 1, p. 282293 (in Russian) [Kur17] Kurkin A. N., Makushkina L. A. Software implementation of methods for automatic ontology generation based on structured material. Scientific journal NovaInfo.ru. No. 52-2, 2017 (in Russian) [Lin16] Lincoln Sheets, Gregory F.Petroski, Yan Zhuang, Michael A.Phinney, Bin Ge, Jerry C.Parker, Chi- Ren Shyu Combining Contrast Mining with Logistic Regression to Predict Healthcare Utilization in a Managed Care Population. Appl Clin Inform. 2017 Apr; 8(2): 430446. Published online 2017 Dec 21. doi: 10.4338/ACI-2016-05-RA-0078 [Lop04] Lopez-Ortega O., Suarez J., Lopez-Morales V. Towards the formalization of knowledge representation. 2004. https://www.researchgate.net/publication/262286370 Towards the formalization of knowledge representati [Mas14] Maslova O.V., Makushkina L.A. Analysis of methods for generating ontological models from a collection of text documents. Bulletin of the magistracy. 2014. No. 4-1 (31). pp. 85 to 89 (in Russian). [Men15] Menshenina I. A. Analysis of the structural features of medical texts. Achievements of fundamental, clinical medicine and pharmacy: materials of the 70th scientific. ses al. University, Jan. 28-29. 2015 - Vitebsk: Voronezh State Medical University, 2015. pp. 336 to 337 (in Russian). 6 [Moz11] Mozherina E.S. Automatic construction of an ontology from a collection of text documents. Proceedings of the 13th All-Russian Scientific Conference Digital Libraries: Advanced Methods and Technologies, Electronic Collections - RCDL2011, Voronezh, Russia, 2011. http://rcdl.ru/doc/2011/paper45.pdf (in Russian). [Muh15] Muhammad Habib ur Rehman, Chee Sun Liew, Teh Ying Wah, Junaid Shuja, Babak Daghighi Mining Personal Data Using Smartphones and Wearable Devices: A Survey. Sensors (Basel). 2015 Feb; 15(2): 44304469. Published online 2015 Feb 13. doi: 10.3390/s150204430 [Mur07] Muromtsev D.I. Ontological knowledge engineering in the Protege system. SPb: SPbSUITMO, 2007. 62 p. (in Russian). [Nes18] Nesrine Zoghlami, Besma Glaa, Souad Rabah, Mourad Abed Healthcare decision support tool: multi- agent system for bed management. Int. J. Applied Management Science, Vol. 10, No. 1, 2018. [Nov18] Novikov D.A. Methods for extracting and analyzing terminological structures of related subject areas (using the methodology as an example). Design Ontology, No. 3 (29), v. 8. 2018. pp. 347-365 [Ord11] Order of the Ministry of Health and Social Development of the Russian Federation of April 28, 2011 N 364 (in Russian) [Orb12] Orbinskaya E. A. A method for automatically constructing a domain ontology based on an analysis of the linguistic characteristics of a text corpus. Proceedings of the XV All- Russian Joint Conference Internet and Modern Society (IMS-2012), St. Petersburg, Russia, 2012, https://arxiv.org/ftp/arxiv/papers/1405/1405.1346.pdf (in Russian) [Pic14] Pichurin V.V. Psychological and psychophysical training as a part of physical education of students in higher educational establishments. Pedagogigs. Psichology, 11: 4448. 2014 [Piv05] A. Pivk Automatic ontology generation from web tabular structures. Doctoral dissertartion. 2005. https://dis.ijs.si/Sandi/docs/PhD-EnglishAbstract.pdf [Pop96] Popov E.V., Fominykh I.B., Kisel E.B., Shapot M.D. Statistical and dynamic expert systems. M., Finance and statistics. 1996 (in Russian) [Rai18] Raikov A. N. Automated synthesis of a cognitive model based on big data analysis and deep learning. Proceedings of the 21st International Joint Conference ”Internet and Modern Society” (St. Petersburg, 2018). SPb .: ITMO University, 2018. Issue 2. P. 103-111 (in Russian) [Ron14] Ronald Margolis, Leslie Derr, Michelle Dunn, Michael Huerta, Jennie Larkin, Jerry Sheehan, Mark Guyer, Eric D. Green The National Institutes of Health’s Big Data to Knowledge (BD2K) initiative: capitalizing on biomedical big data. J Am Med Inform Assoc. 2014 Nov; 21(6): 957958. Published online 2014 Jul 9. doi: 10.1136/amiajnl-2014-002974 [Ros10] Rosales, Romer and Farooq, Faisal and Krishnapuram, Balaji and Yu, Shipeng and Fung, Glenn Au- tomated Identification of Medical Concepts and Assertions in Medical Text. AMIA Annual Symposium proceedings / AMIA Symposium, November 2010, pp. 682-686 [Sai18] Md Saiful Islam, Md Mahmudul Hasan, Xiaoyi Wang, Hayley D.Germack, Md Noor-E-Alam A System- atic Review on Healthcare Analytics: Application and Theoretical Perspective of Data Mining. Health- care (Basel). 2018 Jun; 6(2): 54. Published online 2018 May 23. doi: 10.3390/healthcare6020054 [Sam09] Samkov E.Yu. Representation of knowledge through conceptual semantic networks. Sat tr First All- Russian. scientific and technical conf. ”Systems of organizational behavior (80V’2009).” Voronezh: VGASU, 2009. p. 80-81 (in Russian) [Sha17] K. J.Shakhgeldyan, D. V.Gmar, B. I.Geltser Using Machine learning to Assess efficiency of tuberculosis service. IEEE Xplore Digital Library, 2017, 25-29 sept., DOI: 10.1109/RPC.2017.8168091. Association of Enterprises on Computer and Information Technologies 7 [Sha07] Sharopin K.A. Information system for assessing the psychophysical readiness of students for professional activities. Abstract of dissertation for the degree of candidate of technical sciences, Tomsk Polytechnic University, Tomsk. 2007 (in Russian) [Sha11] Sharopin K.A. Psychophysical readiness of students for professional activities. Assessment Information System. LAP LAMBERT Academic Publishing, pp. 196. 2011 (in Russian) [Sme16] Smetanina O.N., M.M. Gayanova, T.V. Naumova, R.Ch. Gayanov Information aspects of professional applied physical training of students. Proceedings of the International Conference Information Tech- nologies for Intelligent Decision Support, Ufa, May 2016, Vol. 1. pp.186191 (in Russian) [Sme18] Smetanina O.N., T.V. Naumova, A.Yu. Adelmetova, K.V. Nazmieva Formalization of knowledge sup- ported by management decisions. Proceedings of the International Conference Information Technology for Intelligent Decision Support, Ufa, May 2018, Vol. 3. pp. 716 (in Russian) [Tch10] 18. D. Tcharaktchiev, S. Boytcheva, I. Nikolova, E. Paskaleva, G. Angelova, N. Dimitrova Generating Structured Patient Data via Automatic Analysis of Free Medical Text. Global Telemedicine and E- Health. P. 356-359. [Tha19] Thakkar V., Gordon K. Privacy and Policy Implications for Big Data and Health Information Tech- nology for Patients: A Historical and Legal Analysis. Stud Health Technol Inform. 2019. PP.413-417. doi:10.3233/978-1-61499-951-5-413. [TOM] Tomohiro Sawa Leading advances in the utilization of Big Data in the Healthcare industry. White paper Intel Health & Life Sciences, 16 p. [Tra97] Trakhtengerts E.A. Computer analysis in the dynamics of decision making. Devices and control systems. No.1, 1997, p. 49-56 [Ugl17] A. S.Uglov, A. V.Zamyatin Information and software complex for solving personalized medicine prob- lems using data mining. INFORMATION TECHNOLOGIES AND MATHEMATICAL MODELING (ITMM-2017),Proceedings of the XVI International Conference named after A.F. Terpugova. Kazan, September 29 - October 3, 2017, pp. 126-134 (in Russian). [Vas17] Vasiliev S.N. Formalization of knowledge and management based on positively educated languages. Information technology and computing systems. 1/2008. pp. 3-17 (in Russian). [Wit04] V.A. Wittich, G.I. Gusarova, S.I. Kuznetsov, V.V. Pavlov, P.O. Skobelev, O.L. Surnin, L.S. Fedoseeva E.V. Chernov, M.A. Shamashov Network multiagent model of a regions healthcare management system and a system for monitoring the effectiveness and quality of work of doctors in a polyclinic. Healthcare Informatization and compulsory medical insurance systems (www.idmz.ru), No. 11, 2004 (in Russian). [Yat82] Yatsuk V.Ya. Logical-algebraic language with the involvement of A-conversions to represent knowledge in complex technical systems. In: Intelligent Databanks, 1982 (in Russian). [Yus18] Yusupova N.I., Shakhmametova G.R., Mironov V.V., Zulkarneev R.Kh. Statistical and Intelligent Methods of Medical Data Processing. INFORMATION TECHNOLOGY IN INDUSTRY Vol: 6 Issue 2 P. 13-18, 2018 [Yus19] Yusupova N.I., Smetanina O.N., Sazonova E.Yu., Agadullina A.I. Data Mining to support decision- making on improving the psychophysical readiness of a person for successful professional activity. In- formation Technologies and Systems [Electronic resource]: Proceedings of the Seventh Intern. scientific conf., Khanty-Mansiysk, Russia, March 1216, 2019 (IT&S - 2019): scientific. electron. ed. Khanty- Mansiysk, 2019. pp. 126-131 (in Russian). [Zho17] S. Kevin Zhou, Hayit Greenspan, Dinggang Shen Deep Learning for Medical Image Analysis 1st Edition. Academic Press; 1 edition (February 13, 2017) 8