Scholarlink

<p>The adaptive filtering algorithm recursive least squares (RLS) is widely used for the elimination of motion artifacts in electrocardiogram (ECG) signals. However, RLS suffers from the overshoot phenomenon, which affects its filtering performance. In this work, we proposed a data reuse method for ECG signal processing to achieve the suppression of the overshoot phenomenon while filtering the ECG. We investigated the ECG filtering performance of data reuse RLS (DR-RLS) and proposed an improved DR-RLS (iDR-RLS) for ECG filtering by incorporating the morphological features of ECG to interval for data reuse. Compared with the conventional DR-RLS,RLS, VMD, and FIR algorithms, iDR-RLS increased the signal-to-noise ratio of the filtered signals by 4.3221, 1.3132, 6.3114, and 3.2446&#xa0;dB, respectively. The additional computational cost of the proposed method is only 0.08% of the computational cost of the RLS algorithm itself when the data reuse number is 4. In addition, we implemented the proposed algorithm in hardware based on STM32F407 and developed an ECG portable processing system to verify the efficacy of the algorithm. For ECGs with a sample capacity of 5000, the iDR-RLS algorithm exhibits a slightly longer operational duration (0.09&#xa0;s) than the RLS algorithm. In addition, for 36 sample sets from the database, the filtered signal produced by the proposed algorithm demonstrates an error of −0.001060 ± 0.074392 mv compared to the original signal. This study illustrates the efficacy of data reuse in mitigating motion artifact noise in ECG.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Portable Electrocardiogram System Based on iDR-RLS Filtering

  • 期刊论文

  • 发表日期：

    2025年6月19日

查看全文
  • Di Wu,
  • Wen Fan,
  • Jiafeng Ding,
  • Chang Zeng,
  • Long Mo,
  • Zhengchun Liu

摘要

The adaptive filtering algorithm recursive least squares (RLS) is widely used for the elimination of motion artifacts in electrocardiogram (ECG) signals. However, RLS suffers from the overshoot phenomenon, which affects its filtering performance. In this work, we proposed a data reuse method for ECG signal processing to achieve the suppression of the overshoot phenomenon while filtering the ECG. We investigated the ECG filtering performance of data reuse RLS (DR-RLS) and proposed an improved DR-RLS (iDR-RLS) for ECG filtering by incorporating the morphological features of ECG to interval for data reuse. Compared with the conventional DR-RLS,RLS, VMD, and FIR algorithms, iDR-RLS increased the signal-to-noise ratio of the filtered signals by 4.3221, 1.3132, 6.3114, and 3.2446 dB, respectively. The additional computational cost of the proposed method is only 0.08% of the computational cost of the RLS algorithm itself when the data reuse number is 4. In addition, we implemented the proposed algorithm in hardware based on STM32F407 and developed an ECG portable processing system to verify the efficacy of the algorithm. For ECGs with a sample capacity of 5000, the iDR-RLS algorithm exhibits a slightly longer operational duration (0.09 s) than the RLS algorithm. In addition, for 36 sample sets from the database, the filtered signal produced by the proposed algorithm demonstrates an error of −0.001060 ± 0.074392 mv compared to the original signal. This study illustrates the efficacy of data reuse in mitigating motion artifact noise in ECG.