A neural network approach to electrical impedance cardiography

Author

Kumar, Amaresh A R ; Taylor, Bruce C. ; Mulavara, Ajitkumar P. ; Nair, Meera S. ; Gupta, Vineet ; Timmons, William D.

Author_Institution

Dept. of Biomed. Eng., Akron Univ., OH, USA

fYear

1994

fDate

1994

Firstpage

1077

Abstract

A novel technique to determine cardiac stroke volume using electrical impedance variables is presented. Recent studies have shown that stroke volumes calculated using Kubicek and Sramek impedance equations correlate poorly with Doppler ultrasound (R=0.34). However, by applying a neural network to these same impedance variables, we demonstrated greatly improved correlation (R=0.87) and predictability. This non-linear black box approach to impedance cardiography may lead to a more reliable and inexpensive method for the noninvasive measurement of stroke volume and cardiac output

Keywords

cardiology; Doppler ultrasound; cardiac output; cardiac stroke volume; correlation; electrical impedance cardiography; electrical impedance variables; impedance equations; neural network approach; noninvasive measurement; nonlinear black box approach; predictability; Biomedical engineering; Cardiography; Cardiology; Electrodes; Equations; Impedance measurement; Neural networks; Testing; Thorax; Ultrasonic imaging;

fLanguage

English

Publisher

ieee

Conference_Titel

Engineering in Medicine and Biology Society, 1994. Engineering Advances: New Opportunities for Biomedical Engineers. Proceedings of the 16th Annual International Conference of the IEEE

Conference_Location

Baltimore, MD

Print_ISBN

0-7803-2050-6

Type

conf

DOI

10.1109/IEMBS.1994.415320

Filename

415320