FPGA implementation of Ordinal Pattern Analysis algorithm for Early Detection of Cardiovascular diseases

This paper evaluates a low complex and high accuracy signal processing algorithm for early detection of cardiovascular diseases(CVD) and there after it implements the algorithm on a FPGA based System on Chip embedded platform to classify ECG signals. In the proposed algorithm the complexity of the ECG signal is evaluated using conditional entropy of the ordinal patterns. Further the complexity is compared with a threshold to classify the ECG signal. The performance of the algorithm is validated on exhaustive data sets from MIT-BIH data base with diseases like Cardiac Arrhythmia, Junctional Arrhythmia, Premature Junctional Contractions, Ventricular Arrhythmia and heart blocks. The developed algorithm is validated in software platform as well as in the FPGA platform. The prototype is built using XILINX SPARTRAN3 FPGA. The performance of the algorithm and prototype is evaluated in terms of detection rate which is found to be 100.