DOI: 10.25881/BPNMSC.2020.35.43.014


Karpov O.E.1, Penzin O.V.1, Panin I.I.2, 3, Utyashev N.P.1

1 Pirogov National Medical and Surgical Center, Moscow

2 SberMedAI LLC., Moscow

3 Kharkevich Institute for Information Transmission Problems, Moscow


The paper considers an algorithm for detecting focal epileptiform discharges, created using convolutional neural networks.

A negative predictive value, NPV, tending to 100% (97.4–99.7% with a prevalence of 0.1–5%) provides a high likelihood that the site marked by the algorithm as normal does not actually contain epileptiform discharges.

Thus, the algorithm satisfies the initial requirement to search for regions of interest for detailed analysis by a medical specialist.

Keywords: epilepsy, focal discharges, EEG, machine learning, convolutional neural networks.


1. Fisher R, Cross J, French J, et al. Operational classification of seizure types by the International League Against Epilepsy: Position Paper of the ILAE Commission for Classification and Terminology. Epilepsia. 2017; 58(4): 522–530. doi:10.1111/epi.13670.

2. Avakyan GN, Blinov DV, Lebedeva AV, et al. ILAE classification of the epilepsies: the 2017 revision and update. Epilepsia and paroxyzmal conditions. 2017; 9(1): 6–25. (In Russ). doi:10.17749/2077-8333.2017.9.1.006-025.

3. American Clinical Neurophysiology Society. Guideline twelve: guidelines for long-term monitoring for epilepsy. J Clin Neurophysiol. 2008; 25(3): 170–80. doi: 10.1097/WNP.0b013e318175d472.

4. Kemp B, Värri A, Rosa AC, Nielsen KD, Gade J. A simple format for exchange of digitized polygraphic recordings. Electroencephalogr Clin Neurophysiol. 1992; 82(5): 391–393. doi: 10.1016/0013-4694(92)90009-7.

5. American Electroencephalographic Society guidelines for standard electrode position nomenclature. J Clin Neurophysiol. 1991; 8(2): 200–2.

6. LeCun Y, Boser B, Denker JS, et al. Backpropagation Applied to Handwritten Zip Code Recognition, Neural Computation. 1989; 1(4): 541–551.

7. Büyul, A. SPSS: iskusstvo obrabotki informatsii. Analiz statisticheskikh dannykh i vosstanovlenie skrytykh zakonomernostei: Per. s nem. Spb: DiaSoftYuP, 2005. (In Russ).

8. Rebrova OYu. Statisticheskii analiz meditsinskikh dannykh. Primenenie paketa prikladnykh programm STATISTICA. Moscow: MediaSfera, 2002 (In Russ).

9. Fleiss JL, Levin B, Paik MC. Statistical Methods for Rates and Proportions, 3d ed. New Jersey: John Wiley & Sons, 2003.

10. Heston TF. Standardized predictive values. J Magn Reson Imaging. 2014; 39(5): 1338. doi: 10.1002/jmri.24564.

11. Altman DG, Machin D, Bryant TN, et al. Statistics with Confidence, 2nd ed. UK: British Medical Journal Books; 2000.

12. Penzin OV, Shvyrev SL, Zarubina TV. Results of implementation in the clinical practice the prognostic model for assessing the risk development of mielotoxic complications of chemotherapy. Vestnik novykh medicinskikh tekhnologij. 2019; 1: 112–118. (In Russ). doi: 10.24411/1609-2163-2019-16061.

For citation

Karpov O.E., Penzin O.V., Panin I.I., Utyashev N.P. Analysis of EEG signals using machine learning technologies for detecting focal epileptiform discharges. Bulletin of Pirogov National Medical & Surgical Center. 2020;15(4):69-73. (In Russ.)