Real time cardiac arrhythmia detection

Abstract EN

This paper proposes a real-time implementation of arrhythmia détection on FPGA.

An adaptation ofthe Pan and Tompkins algorithm is made.

It is performed in three main phases, the first resides in the modification of the non-linear transform of the algorithm, by replacing it with the absolute value.

The second in the intégration was achieved by summing 5 elements of a moving window and finally the third derive, validate, count the number ofpeaks to finally make the décision as to the diagnosis.

Using the Analog Descovery device and the XSG design tool, we have realized a hardware cardiac arrhythmia détection system.

According to the simulation results, the methodgives a 96 % accuracy rate, which reveal a good performance of the developed architecture.

A dataset of MIT-BIH arrhythmia database is used in simulation and test of the arrhythmia détection.

This work was implemented using the minimum resources ofthe Vertix5FPGA board.