Voronezh, Voronezh, Russian Federation
employee
Voronezh, Voronezh, Russian Federation
The paper presents a solution to the problem of developing an artificial neural network for classifying electrocardiograms in order to identify a patient's cardiovascular diseases. A literary review of the use of artificial neural networks for use in medicine in the determination of various cardiac pathologies in the condition of patients has been carried out. The most significant six parameters of the heart rhythm have been determined, which can be used in neural network modeling of the state of the human cardiovascular system. An analysis of the capabilities of well-known commercial programs for the classification of electrocardiograms was carried out, their advantages and disadvantages were identified. The capabilities of modern devices for taking cardiograms are described with an indication of the accuracy of their measurements. A study of three different structures of the desired neural network was carried out and the best option was chosen. The software package is made in Python in the Pycharm 2023.1 environment. The network was trained on the basis of the error backpropagation algorithm, the Adam method served as an optimizer. The standard deviation of the network output signal from the standard was used as a criterion. The Softmax function used in the output layer is used to convert the output of neurons into probabilities. Thus, a neural mathematical model for detecting heart diseases has been developed, which has low system requirements for a computer, which can be used to develop software for wearable electronics. An artificial neural network has the following metric: accuracy - 90%, sensitivity - 90%, specificity - 10%.
Artificial intelligence, neural networks, machine learning, ANN learning methods, ECG classification ANN, cardiovascular diseases.
1. Kireeva, K.A. Analiz vozmozhnosti primeneniya iskusstvennyh neyrosetey v kardiologii / K.A. Kireeva, L.A. Korobova, D.V. Arapov // Informacionnye tehnologii v stroitel'nyh, social'nyh i ekonomicheskih sistemah. - 2022. - № 4(30). - S. 168-176.
2. Halaydzhi, A.K. Metody klassifikacii narusheniy serdechnogo ritma na osnove kodirovaniya posledovatel'nostey RR-intervalov signala EKG / A.K. Halaydzhi, I.B. Muchnik // Trudy NGTU im. R.E. Alekseeva. - 2021. - № 1(132). - S. 38-53. - DOI:https://doi.org/10.46960/1816-210X_2021_1_38.
3. Diagnostika gipertrofiy levyh otdelov serdca s pomosch'yu glubokoy neyronnoy seti / P.K. Andreev, V.V. Anan'ev, V.A. Makarov [i dr.] // Trudy Instituta sistemnogo programmirovaniya RAN. - 2020. - T. 32, № 4. - S. 141-154. - DOI:https://doi.org/10.15514/ISPRAS-2020-32(4)-10.
4. Neyrosetevaya model' dlya ranney diagnostiki hronicheskoy serdechnoy nedostatochnosti / V.I. Gorbachenko, V.V. Potapov, O.K. Zenin [i dr.] // Izvestiya vysshih uchebnyh zavedeniy. Povolzhskiy region. Medicinskie nauki. - 2022. - № 4(64). - S. 5-15. - DOI:https://doi.org/10.21685/2072-3032-2022-4-1.
5. Kirpikov, M.V. Metody obucheniya INS / M.V. Kirpikov, A.E. Nemovlenko, A.V. Makarova // Herald of Science and Education. - 2021. - № 11-2(114). - S. 56-60.
6. Onischenko, P.S. Iskusstvennye neyronnye seti v kardiologii: analiz graficheskih dannyh / P.S. Onischenko, K.Yu. Klyshnikov, E.A. Ovcharenko // Byulleten' sibirskoy mediciny. - 2021. - T. 20, № 4. - S. 193-204. - DOI:https://doi.org/10.20538/1682-0363-2021-4-193-204.
7. Shvec, D.A. Sravnitel'nyy obzor ispol'zovaniya metodov mashinnogo obucheniya dlya prognozirovaniya serdechno-sosudistogo riska / D.A. Shvec, S.V. Povetkin // Vestnik novyh medicinskih tehnologiy. Elektronnoe izdanie. - 2020. - № 5. - S. 74-82. - DOI:https://doi.org/10.24411/2075-4094-2020-16711.
8. Ryazanova, S.V. Novye tendencii stanovleniya iskusstvennogo intellekta v medicine / S.V. Ryazanova, V.P. Mazaev, A.A. Komkov // CardioSomatika. - 2021. - T. 12, № 4. - S. 227-233. - DOI:https://doi.org/10.17816/22217185.2021.4.201264.
9. Bacina, E.A. Cifrovizaciya zdravoohraneniya RF: mif ili real'nost'? / E.A. Bacina, A.N. Popsuyko, G.V. Artamonova // Vrach i informacionnye tehnologii - 2020 - № 3. S. 73-80. - DOI:https://doi.org/10.37690/1811-0193-2020-3-73-80.
10. Smartfon v medicine - ot spravochnika k diagnosticheskoy sisteme. Obzor sovremennogo sostoyaniya voprosa / A.A. Fedorovich, A.Yu. Gorshkov, A.I. Korolev, O.M. Drapkina // Kardiovaskulyarnaya terapiya i profilaktika. - 2022. - № 21(9). - S. 3298. - DOI:https://doi.org/10.15829/1728-8800-2022-3298.
11. Biblioteka glubokogo obuchenie na Python. - URL: https://ru-keras.com/(data obrascheniya: 10.04.2023).
12. Baza dannyh MIT-BIH Arrythmia Database. - URL: https://physionet.org/content/ptbdb/1.0.0/(data obrascheniya: 10.04.2023).
13. Pieszko, K. Predicting Long-Term Mortality after Acute Coronary Syndrome Using Machine Learning Techniques and Hematological Markers / K. Pieszko, J. Hiczkiewicz, P. Budzianowski // Dis Markers. - 2019. - Vol. 1. - C. 9056402. - DOI:https://doi.org/10.1155/2019/9056402.
14. A survey on ECG analysis / S.K. Berkaya, A.K. Uysal, E.S. Gunal [et al.] // Biomedical Signal Processing and Control. - 2018. - Vol. 43. - Pp. 216-235. - DOI:https://doi.org/10.1016/j.bspc.2018.03.003.
15. Fractional flow reserve or coronary flow reserve for the assessment of myocardial perfusion: Implications of FFR as an imperfect reference standard for myocardial ischemia / V.E. Stegehuis, G.W. Wijntjens, J.J. Piek, T.P. van de Hoef // Current Cardiology Reports. - 2018. - Vol. 20(9). - C. 77. - DOI:https://doi.org/10.1007/s11886-018- 1017-4.
16. Deep learning analysis of left ventricular myocardium in CT angiographic intermediate-degree coronary stenosis improves the diagnostic accuracy for identification of functionally significant stenosis / R.W. van Hamersvelt, M. Zreik, M. Voskuil [et al.] // European Radiology. - 2019. - Vol. 29(5). - Pp. 2350-2359. - DOI:https://doi.org/10.1007/s00330-018-5822-3.
17. Catheter localization in 3D ultrasound using voxel-of-interest-based ConvNets for cardiac intervention / H. Yang, C. Shan, A.F. Kolen, P.H.N. de With // International Journal of Computer Assisted Radiology and Surgery. -2019. - Vol. 14(6). - Pp. 1069-1077. - DOI:https://doi.org/10.1007/s11548-019- 01960-y.
18. Deep learning-based muscle segmentation and quantification at abdominal CT: Application to a longitudinal adult screening cohort for sarcopenia assessment / P.M. Graffy, J. Liu, P.J. Pickhardt [et al.] // The British Journal of Radiology. - 2019. - Vol. 92(1100). - Pp. 2921-2928. - DOI:https://doi.org/10.1259/bjr.2019032.
19. Full left ventricle quantification via deep multitask relationships learning / W. Xue, G. Brahm, S. Pandey [et al.] // Medical Image Analysis. - 2018. - Vol. 43. - Pp. 54-65. -DOI:https://doi.org/10.1016/j.media.2017.09.00.
20. Zhou, T. A review: Deep learning for medical image segmentation using multi-modality fusion / T. Zhou, S. Ruan, S. Canu // Array. - 2019. - Vol. 3-4. - C. 100004. - DOI:https://doi.org/10.1016/j.array.2019.100004.
21. Deep learning in medical imaging / M. Kim, J. Yun, Y. Cho [et al.] // Neurospine. - 2019. -Vol. 16(4). - Pp. 657-668. - DOI:https://doi.org/10.14245/ns.1938396.198.
22. Pesapane, F. Artificial intelligence in medical imaging: threat or opportunity? Radiologists again at the forefront of innovation in medicine / F. Pesapane, M. Codari, F. Sardanelli // European Radiology Experimental. - 2018. - Vol. 2(1). - C. 35. - DOI:https://doi.org/10.1186/s41747-018-0061-6.
23. Brain tumor segmentation using convolutional neural networks in MRI images / M.M. Thaha, K.P.M. Kumar, B.S. Murugan [et al.] // Journal of Medical. Systems. - 2019. - Vol. 43(9). - Pp. 1240-1251. - DOI:https://doi.org/10.1007/s10916-019-1416-0.
24. Automated cardiovascular magnetic resonance image analysis with fully convolutional networks / W. Bai, M. Sinclair, G. Tarroni [et al.] // Journal of Cardiovascular Magnetic Resonance. - 2018. -Vol. 20(1). - C. 65. - DOI:https://doi.org/10.1186/s12968-018-0471-x.
25. Saymon, D. Karmannyy spravochnik po EKG / D. Saymon. - M. : GEOTAR Media Rossiya, 2020. - 192 s.
26. Osnovy elektricheskoy elektrokardiografii : uchebnoe posobie // I.G. Men'shikova, E.V. Magalyas, I.V. Sklyar, N.V. Loskutova. - Blagoveschensk, 2021. - 112 s.
27. Elektrokardiografy (EKG apparaty). - URL: https://permedcom.ru/catalog/funktsionalnaya-diagnostika/elektrokardiografy/(data obrascheniya: 10.04.2023).
28. Ekg-apparaty. - URL: https://lpu.elamed.com/katalog/medicinskoe-oborudovanie/diagnosticheskoe-oborudovanie/ekg-apparaty/(data obrascheniya: 10.04.2023).
29. Program Development for Choosing a Surgical Treatment Option and Mathematical Prediction of Findings in Patients with Postoperative Median Abdominal Hernias / D.V. Arapov, E.F. Cherednikov, S.A. Skorobogatov [et al.] // International Journal of Biomedicine. - 2022. - Vol. 12, No. 2. - P. 303-307. - DOI:https://doi.org/10.21103/Article12(2)_OA19.
30. Information system for diagnosis of respiratory system diseases / G.V. Abramov, L.A. Korobova, A.L. Ivashin, I.A. Matytsina // Journal of Physics: Conference Series. - 2018. - Vol. 1015, 4. -P. 042036. - DOI:https://doi.org/10.1088/1742-6596/1015/4/042036.
31. Razrabotka modulya diagnostiki informacionnoy sistemy monitoringa zdorov'ya bol'nyh legochnymi zabolevaniyami / G.V. Abramov, L.A. Korobova, A.L. Ivashin, I.A. Matycina // Vestnik Ryazanskogo gosudarstvennogo radiotehnicheskogo universiteta. - 2018. - № 65. - S. 136-142. - DOI:https://doi.org/10.21667/1995-4565-2018-65-3-136-142.
32. Korobova, L.A. Razrabotka medicinskoy ekspertnoy sistemy diagnostiki zabolevaniy s ispol'zovaniem elementov teorii mnozhestv / L.A. Korobova, E.S. Malienko, Yu.A. Safonova // Ekonomika i menedzhment sistem upravleniya. - 2017. - № 4-1(26). - S. 172-178.
33. Korobova, L.A. Razrabotka modeli prinyatiya resheniya dlya postanovki diagnoza zabolevaniy na osnove nechetkoy logiki / L.A. Korobova, T.V. Gladkih // Vestnik Voronezhskogo gosudarstvennogo universiteta inzhenernyh tehnologiy. - 2018. - T. 80, № 4(78). - S. 80-89. - DOI:https://doi.org/10.20914/2310-1202-2018-4-80-89.
34. Korobova, L.A. Razrabotka modulya medicinskoy informacionnoy sistemy dlya diagnostiki legochnyh zabolevaniy / L.A. Korobova, I.A. Matycina // Vestnik NGIEI. - 2018. - № 10(89). - S. 13-28.
35. Prototype mobile application definitions fresh products based on neural network / L.A. Korobova, I.S. Tolstova, I.A. Matytsina, M.S. Mironova // Journal of Physics: Conference Series. - 2021. - Vol. 1902. - P. 012118. - DOI:https://doi.org/10.1088/1742-6596/1902/1/012118.
