{"title":"Determination of the Signal-to-Noise Ratio for Noisy Electrocardiogram Using Lossless Data Compression","authors":"A. Pulavskyi, S. Krivenko, L. Kryvenko","doi":"10.1109/MECO.2019.8760294","DOIUrl":null,"url":null,"abstract":"A numerical method for determining the signal-to-noise ratio (SNR) in electrocardiograms (ECG), distorted by various types of additive noise, is proposed. It is based on the calculation of the compression ratio (CRn) of an ECG fragment of 10 seconds duration. The functional dependences of SNR on CRn for ECG distorted by electrode motion, muscle artefacts and mixed noise, were found, and they are exponential. The values of root mean square error (RMSE) for such dependencies were 4.24 dB, 3.57 dB и 3.33 dB accordingly for SNR in range −10…45 dB. It is proved that the inclusion of heart rate (HR) in the model reduces the error by 0.6-0.8 dB. The present approach is based on the standard library zlib and is recommended for use in portative mobile systems.","PeriodicalId":141324,"journal":{"name":"2019 8th Mediterranean Conference on Embedded Computing (MECO)","volume":"52 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2019-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2019 8th Mediterranean Conference on Embedded Computing (MECO)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/MECO.2019.8760294","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
A numerical method for determining the signal-to-noise ratio (SNR) in electrocardiograms (ECG), distorted by various types of additive noise, is proposed. It is based on the calculation of the compression ratio (CRn) of an ECG fragment of 10 seconds duration. The functional dependences of SNR on CRn for ECG distorted by electrode motion, muscle artefacts and mixed noise, were found, and they are exponential. The values of root mean square error (RMSE) for such dependencies were 4.24 dB, 3.57 dB и 3.33 dB accordingly for SNR in range −10…45 dB. It is proved that the inclusion of heart rate (HR) in the model reduces the error by 0.6-0.8 dB. The present approach is based on the standard library zlib and is recommended for use in portative mobile systems.
京公网安备 11010802042870号
京ICP备2023020795号-1
分享
求助内容：
应助结果提醒方式：
扫码关注我们
