V.I. Zhulev1, A.V. Kroshilin2, S.V. Kroshilina3

1–3 Ryazan State Radio Engineering University named after V.F. Utkina (Ryazan, Russia)

The problems associated with the presentation of knowledge in medical subject areas require elaboration in the automation of medical decision support. There are a large number of different approaches to the creation of intelligent systems and to the presentation of artificial intelligence tasks, among which two main ones can be noted. The first approach is called top-down (semiotic), it consists in creating all kinds of expert systems, knowledge bases and logical inference systems that allow simulating various high-level mental processes: speech, reasoning, thinking, creativity, emotions and the like. The second approach, bottom-up (biological), consists in the study of various evolutionary calculations and neural networks that build a model of intellectual behavior and are based on smaller and "non-intellectual" elements.

The purpose of the work is to determine and justify the presentation of medical knowledge on the basis of linguistic variables that should take into account the peculiarities of the medical subject area. Give a description of the fuzzy system and the additional requirements, as well as tables of statements that are fuzzy. To present the general structure of medical expert systems, identify the features and consider the main functions. To review expert systems, including medical expert systems.

The analysis of the features of knowledge formation in managerial medical solutions, as well as the problems arising during their implementation in intelligent information systems, is carried out. It is shown that in order to implement fuzzy knowledge models, it is necessary to choose tools for their construction and work with them (formal logical systems, linguistic variables, fuzzy variables, fuzzy systems, fuzzy logical rules and basic schemes of fuzzy inference). The main types of models for building medical knowledge are given, and it is concluded that most of the models that allow obtaining new knowledge based on existing ones can be reduced to a productive language for building knowledge. The features of knowledge formation in medical systems are considered. A number of features and functions of medical expert systems are listed, a general structure is developed.

The applied intelligent data analysis systems for medical decision-making were analyzed, and an example of the structure of a medical expert system was given. A fuzzy inference approach is proposed to formalize medical experience. The scheme of the process of acquiring knowledge in the information system of medical decision support is presented. An overview of expert systems and medical expert systems is given. A feature of medical expert systems is working with one of two types of databases (empirical or medical). A hybrid construction scheme is required that allows you to work with two types of knowledge simultaneously.

Zhulev V.I., Kroshilin A.V., Kroshilina S.V. Knowledge formation and structure of the medical expert system. Biomedicine Radioengineering. 2023. V. 26. № 3. Р. 44-54. DOI: https://doi.org/10.18127/j15604136-202303-06 (In Russian).

1. Kroshilina S.V., Kroshilin A.V., Zhuleva S.Yu. Predstavleniye znaniy na osnove teorii nechetkikh mnozhestv v meditsinskikh predmetnykh oblastyakh. Biomeditsinskaya radioelektronika. 2022. T. 25. № 4. S. 62–70. DOI: https://doi.org/10.18127/j15604136-202204-08 (in Russian)
2. Tishkina V.V., Kroshilin A.V., Pylkin A.N. Application of Fuzzy Logic in Decision Support System for Analysis of Condition Enterprises. 2018 Internation-al Russian Automation Conference (RusAutoCon 2018). September 9th-16th. 2018. Sochi. Russian Federation (Primeneniye nechetkoy logiki v sisteme pod-derzhki prinyatiya resheniy dlya analiza sostoyaniya predpriyatiy). (in Russian).
3. Pylkin A.N., Kroshilin A.V., Kroshilina S.V. Algoritm modifitsi-rovannogo metoda nechetkoy klasterizatsii v intellektualnykh meditsinskikh sistemakh. Matematicheskiye i kompyuternyye metody v meditsine. biologii i ekologii: monografiya. Pod nauch. red. V.I. Levina. Vyp.2. S. 54–65. Penza; Moskva: Privolzhskiy Dom znaniy; MIEMP. 2013. 112 s. (in Russian).
4. Zhuleva S.Yu., Kroshilin A.V., Kroshilina S.V. Podderzhka prinyatiya upravlencheskikh meditsinskikh resheniy i priroda neopredelennosti v nikh. Biomeditsinskaya radioelektronika. 2021. T. 24. № 4. S. 89-96. (in Russian).
5. Kroshilin A., Kroshilina S., Pylkin A., Ovechkin G. Managerial medical decisions and methods of obtaining medical information in conditionsof uncertain-ty. 2021 10th Mediterranean Conference On Embedded Computing (Meco2021). 7-10 June 2021. Budva. Montenegro. 864 p. P. 500–503.
6. Alekseyev A.V. Intellektualnyye sistemy prinyatiya proyektnykh resheniy. A.V. Alekseyev, A.N. Borisov, E.R. Vilyums, N.N. Slyadz, S.A. Fomin. Riga: Zinatne. 1997. 320 s. (in Russian).
7. Zade L. Ponyatiye lingvisticheskoy peremennoy i eye primeneniye k prinyatiyu priblizhennykh resheniy: per. s angl. N.I. Ringo. M.: Mir. 1976. 168 s. (in Russian).
8. Minto V. Deduktivnaya i induktivnaya logika. Ekaterinburg: Delovaya kniga; – Bishkek: Odissey. 1997. 432 s. (in Russian).
9. Kroshilin A.V., Kroshilina S.V., Zhuleva S.Yu. Realizatsiya modeli dinamiki raspredeleniya materialnykh potokov v meditsinskom uchrezhdenii. Biomeditsinskaya radioelektronika. 2020. T. 23. № 3. S. 53–60. (in Russian).
10. Gavrilova T.A., Khoroshevskiy V.F. Bazy znaniy intellektualnykh sistem. SPb.: Piter. 2001. 384 s. (in Russian).
11. Osipov G.S. Priobreteniye znaniy intellektualnymi sistemami. M.: Nauka. Fizmatlit. 1997. 112 s. (in Russian).
12. Kroshilin A.V., Kroshilina S.V. Obzor sushchestvuyushchikh meditsinskikh ekspertnykh sistem. Matematicheskoye i programmnoye obespecheniye vychisli-telnykh sistem: Mezhvuz. sb. nauch. tr. Ryazan: RGRTU. 2010. S. 122–125. (in Russian).