DocumentCode :
406551
Title :
A novel neural model of cardiac action potential
Author :
Guerreiro, Ana M G ; de Araujo, Carlos A Paz
Author_Institution :
Dept. of Electr. & Comput. Eng., Colorado Univ., Colorado Springs, CO, USA
Volume :
2
fYear :
2003
fDate :
17-21 Sept. 2003
Firstpage :
1893
Abstract :
This paper presents a novel neuron model for information processing in neural networks of the spike-neuron class. The model called hybrid biological neuron, HBN is trained to be an autonomous oscillator, thus is able to model cardiac action potentials and frequency of cardiac rhythmicity. The HBN is a new model that allows pulse trains to be interpreted on one hand as a logic function, and on the other hand as a continuous time system in which the pulse shape represents a second order modulation of the information not encoded in the patterns only. The HBN belongs to a spiking neuron class but models chemical synapses and considers the receptive field caused by the direct influence of other dendrites, specifically the ones that do not pass through the pre-synaptic filters.
Keywords :
bioelectric potentials; cardiology; continuous time systems; learning (artificial intelligence); muscle; neural nets; neurophysiology; oscillators; physiological models; autonomous oscillator; cardiac action potential; cardiac rhythm frequency; chemical synapses; continuous time system; dendrites; hybrid biological neural model; information processing; neural networks; pre-synaptic filters; pulse shape; pulse trains; receptive field; second order modulation; spike-neuron class; Biological information theory; Biological system modeling; Frequency; Information processing; Logic functions; Neural networks; Neurons; Oscillators; Pulse modulation; Pulse shaping methods;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Engineering in Medicine and Biology Society, 2003. Proceedings of the 25th Annual International Conference of the IEEE
ISSN :
1094-687X
Print_ISBN :
0-7803-7789-3
Type :
conf
DOI :
10.1109/IEMBS.2003.1279789
Filename :
1279789
Link To Document :
بازگشت