Share Email Print
cover

Proceedings Paper • new

Convolutional neural networks application in cardiovascular decision support systems
Author(s): Natalia Konnova; Mikhail Basarab; Michael Khachatryan; Anna Domracheva; Igor Ivanov
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

The paper considers the possibilities of using neural network methods of machine learning to diagnose the states of the human cardiovascular system and support decision-making in cardiology and cardiac surgery. The issues of processing and preparation of electrocardiography signals, selection of architecture and tuning of neural network parameters for automation of diagnosis are discussed. Here, the results obtained with the help of multilayer perceptrons and convolutional neural networks to assign the submitted input cardiovascular data to one of the classes of states in the selected space are examined. Based on a specialized developed software, the proprietary numerical experiments with real clinical data were carried out. Given the above results, demonstrating the applicability of the used deep learning methods and algorithms to diagnostic automation, a model of a hierarchical decision support system is proposed.

Paper Details

Date Published: 27 November 2019
PDF: 6 pages
Proc. SPIE 11321, 2019 International Conference on Image and Video Processing, and Artificial Intelligence, 113212D (27 November 2019); doi: 10.1117/12.2548193
Show Author Affiliations
Natalia Konnova, Bauman Moscow State Technical Univ. (Russian Federation)
Mikhail Basarab, Bauman Moscow State Technical Univ. (Russian Federation)
Michael Khachatryan, Bauman Moscow State Technical Univ. (Russian Federation)
Anna Domracheva, Bauman Moscow State Technical Univ. (Russian Federation)
Igor Ivanov, Bauman Moscow State Technical Univ. (Russian Federation)


Published in SPIE Proceedings Vol. 11321:
2019 International Conference on Image and Video Processing, and Artificial Intelligence
Ruidan Su, Editor(s)

© SPIE. Terms of Use
Back to Top