Patients Identification in Medical Organization Using Neural Network

  • V. Evdokimov Alexey
  • A. Kovalenko Vasiliy
  • V. Kurbesov AlexandrEmail author
  • V. Shabanov Alexey
Conference paper
Part of the Advances in Intelligent Systems and Computing book series (AISC, volume 896)


This article describes the method of patients identification by using neural network. Our complex filters increases recognition quality and greatly reduces computing resources needed for successful identification of a patient and can be applied to any neural network, used for identification. The resulting software is developed using only open source components and needs about 1 s to recognize a patient. Our experiment showed the effectiveness of recognition of patients not less than 99.9% with false positive ratio less than 0.00001 when using the software at the medical organization. Time interval was calculated from the first face recognition in video stream to successful identification with Euclidean metric less than 0.4. Frame rate of video stream was 25 frames per second. We measured the speed of the doctors’ work before and after the introduction of the biometric identification system for patients. The patient’s reception time is reduced by an average of 5 s. Based on 20.000 visits per month and 7 min per person on average, it is possible to take an additional 238 people a month without increasing the number of doctors.


Optimization Neural network Identification Image quality estimation 


  1. 1.
    Marichal, X., Wei-Ying, M., HongJiang, Z.: Blur determination in the compressed domain using DCT information. In: Proceedings of 1999 International Conference on Image Processing, ICIP 1999, vol. 2 (1999)Google Scholar
  2. 2.
    Kalalembang, E., Usman, K., Gunawan, I.P.: DCT-based local motion blur detection. In: International Conference o Instrumentation, Communications, Information Technology, and Biomedical Engineering (ICICI-BME) (2009)Google Scholar
  3. 3.
    Huynh-Thu, Q., Ghanbari, M.: The accuracy of PSNR in predicting video quality for different video scenes and frame rates. Telecommun. Syst. 49(1), 35–48 (2012)CrossRefGoogle Scholar
  4. 4.
    Schroff, F., Kalenichenko, D., Philbin, J.: FaceNet: A unified embedding for face recognition and clustering. In: Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition (2015)Google Scholar
  5. 5.
    Chen, D., Ren, S., Wei, Y., Cao X., J. Sun J.: Joint cascade face detection and alignment. In: Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition (2014)Google Scholar
  6. 6.
    Weinberger, K.Q., Blitzer, J., Saul, L.K.: Distance metric learning for large margin nearest neighbor classification, pp. 2–3. In: NIPS. MIT Press (2006)Google Scholar
  7. 7.
    Wilson, D.R., Martinez, T.R.: The general inefficiency of batch training for gradient descent learning. Neural Netw. 16(10), 1429–1451 (2003)CrossRefGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  • V. Evdokimov Alexey
    • 1
  • A. Kovalenko Vasiliy
    • 1
  • V. Kurbesov Alexandr
    • 2
    Email author
  • V. Shabanov Alexey
    • 1
  1. 1.Electronic Medicine Ltd.Rostov-on-DonRussia
  2. 2.Rostov State University of EconomicsRostov-on-DonRussia

Personalised recommendations