PDF

Abstract:
The article discusses results of the development of an expert system module for diagnosing diseases based on the method of neural network analysis. In the course of the study, it was established that when processing images obtained using magnetic resonance imaging (MRI) devices, convolutional neural networks offer the maximum efficiency. An algorithm is developed to select an optimal neural network structure best suited for our objective. As a result of the work, we developed a convolutional neural network capable of detecting foci of pathological changes in tissues with high probability in the images obtained by MRI scanners. The method was evaluated using a separate human organ - lungs. The system was implemented in a test mode in one of the largest clinics of the city of Voronezh.

Keywords:
diagnostics, binarization, clustering, classification, convolutional neural network.

Citation:
Vasilchenko VA, Burkovskiy VL, Danilov AD. Algorithmization of the process of recognition of states of living objects based on special x-ray images. Computer Optics 2019; 43(2): 296-303. DOI: 10.18287/2412-6179-2017-43-2-296-303.

References:

1. Magnetic resonance imaging. 2018. Source: < https://en.wikipedia.org/wiki/Magnetic_resonance_imaging >.
2. Vasilchenko VA, Burkovskiy VL. Analysis and choice of the structure of the object-relational DBMS of the medical information system [In Russian]. Bulletin of Voronezh State Technical University 2015; 3(3): 57-59.
3. Vasilchenko VA, Burkovskiy VL. Structure of the expert system of operative diagnosis and treatment of lung diseases [In Russian]. Bulletin of Voronezh State Technical University 2016; 12(1): 28-31.
4. Maksimova EI, Khaustov PA. Algorithm for detecting formations in the human lungs on images of a computer tomograph using an artificial neural network [In Russian]. Basic Research 2016; 4(2): 290-294.
5. EmguCV. 2017. Source: áhttp://www.emgu.com/wiki/ index.php/Main_Pageñ.
6. Canny J. A computational approach to edge detection. IEEE Transactions on Pattern Analysis and Machine Intelligence. 1986; 6: 679-698.
7. Kruglov VI, Dli MI, Golunov RY. Fuzzy logic and artificial neural networks [In Russian]. Moscow: "Fizmatlit" Publisher; 2002.
8. Gonzalez RC, Woods RE. Digital image processing. Upper Saddle River, NJ: Prentice-Hall, Inc; 2006.
9. Criffith AK. Edge detection in simple scenes using a priori information. IEEE Trans on Computers 1971; 5: 551-561.
10. How to choose algorithms for Azure Machine Learning Studio. Source: < https://docs.microsoft.com/en-us/azure/ machine-learning/studio/algorithm-choice >.
11. ALGLIB. Source: < http://www.alglib.net >.
12. Sprindzhuk MV, Kovalev VA, Snezhko EV, Dmitruk AA,Bogush AL, Khoruzhik SA. Computer-assisted diagnosis of nodal structures in the lungs (literature review) [In Russian]. Povolzhsky oncologybulletin 2010. Source: < http://oncovestnik.ru/index.php/diagnostika/item/864-kompyuter-assistirovannaya-diagnostika-uzlovykh-obrazovanij-v-legkikh-obzor-literatury >.
13. Mistyukov BV, Burkovskiy VL. Neural network modeling of the dynamics of nonlinear systems [In Russian]. Bulletin of Voronezh State Technical University 2012; 3(12-1): 51-56.
14. Goncharova YM, Pyanikov AE, BurkovskiyVL.Neural network modeling of acute pyelonephritis in the experiment [In Russian]. Bulletin of Voronezh State Technical University 2012; 8(7-1): 54-57.
15. Spirayachin AA, Burkovskiy VL, Voropayev AP. Analysis of the effectiveness of the use of the fuzzy classification method and genetic algorithms in the intellectual decision support system of medical decisions [In Russian]. Bulletin of Voronezh State Technical University 2015; 11(2): 45-48.

---

© 2009, IPSI RAS
151, Molodogvardeiskaya str., Samara, 443001, Russia; E-mail: journal@computeroptics.ru ; Tel: +7 (846) 242-41-24 (Executive secretary), +7 (846)332-56-22 (Issuing editor), Fax: +7 (846) 332-56-20