Vol. 64 | nr. 1 | Ianuarie - Martie 2016 Articolul 20

Implementation of a Neural Classifier on a FPGA-based Reconfigurable System of Cardiac Arrhythmias
(Full text in English)

Amel BABA HAMED, Hassane BECHAR, Mohammed Amine CHIKH
Abstract
This work proposes a hardware method for cardiac arrhythmias classification, the neural classifier was implemented on a FPGA programmable circuit using a high level language of hardware description VHDL. The parameters of the neural classifier are first calculated on offline. Then they are implemented in an FPGA in order to have a realtime neural classifier. The hardware neural classifier is able to detect overflows. The neural classifier was tested using the simulation tool ModelSim SE6.2c, and then synthesized under the Xilinx ISE14.7 Design environment.
Keywords: Neural networks, Multilayer Perceptron, VHDL, FPGA, CAD tools, Real Time, Cardiac Arrhythmias

 

Download