Hardware/Software Co-design of an ECG- PPG Preprocessor: A Qualitative & Quantitative Analysis

Abstract

This paper aims to design a digital system to pre-process electrocardiogram (ECG) and photoplethysmogram (PPG) signal for the purpose of hardware implementation. Muscle signal, motion artifacts, power line interference affect the biomedical signal during data acquisition. The proposed system focuses at removing the noises by designing infinite impulse response filter to remove power line noise and finite impulse response filter to eliminate other high and low frequency noises. At first the preprocessor is designed in Matlab to validate the simulation performance. Then the hardware is designed in xilinx system generator targeting Zedboard Zynq xc7z020-1clg484. Finally, we verified the hardware software codesign by comparing both outputs. For quantity based analysis different filtering techniques have been applied to determine the most optimized system in terms of resource utilization and power consumption. Pearson correlation coefficient of 0.9993 and 0.9982 have been found for ECG and PPG, respectively using Hamming filter technique for High and low pass filter. Root squared error for both signal has been also in the range of 10−2• These data validate the accuracy of the designed system providing quality assurance. Frequency spectrum also has been analyzed to ensure denoising of undesired signals. The designed preprocessor can be utilized for further analysis of the signals and designing digital systems & wearable devices for the detection of heart rate, cardiac diseases etc.