Scholarlink

<p>Recording the electrical activity of the heart is done using an electrocardiogram (ECG). Compressed sensing's classification skills have recently been used to cardiovascular disease monitoring, allowing for more efficient patient monitoring using compressed physiological data. The quantity of data produced by the sensor depends on how long it has been monitoring the signal. Reducing storage space is possible with the use of an efficient lossless ECG compression technique. A hardware design for a multiple channels lossless ECG compression method is shown in this short. The algorithm that forms the basis of the system incorporates both adaptive linear prediction (ALP) and multi-channel linear prediction (MLP). For entropy coding, the Golomb rice coder (GRC) is also used. Optimal hardware resources were used in the design of the hardware implementation, which aimed to minimize hardware complexity utilization. To top it all off, the design allows for great throughput by processing several channels simultaneously. Therefore, instead of processing individual heartbeats, arrhythmia categorization from compressed ECG data must be done in segments of predetermined length. Following this, we provide a deep learning (DL) model that, with the benefits of a high compression ratio (CR) and a low processing cost, can directly identify various forms of arrhythmia using compressed ECG segments of a set length. Our suggested strategy achieves an exact match rate of 97.03% at CR(Compression Ratio), according to experimental findings on the MITBIH arrhythmias database.</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

Low power negative capacitance field effect transistor (NCFET) design for ECG applications using compression and A-CNN based classification

  • 期刊论文

  • 发表日期：

    2025年11月19日

查看全文
  • E. Ramya,
  • N. Kumaresan

摘要

Recording the electrical activity of the heart is done using an electrocardiogram (ECG). Compressed sensing's classification skills have recently been used to cardiovascular disease monitoring, allowing for more efficient patient monitoring using compressed physiological data. The quantity of data produced by the sensor depends on how long it has been monitoring the signal. Reducing storage space is possible with the use of an efficient lossless ECG compression technique. A hardware design for a multiple channels lossless ECG compression method is shown in this short. The algorithm that forms the basis of the system incorporates both adaptive linear prediction (ALP) and multi-channel linear prediction (MLP). For entropy coding, the Golomb rice coder (GRC) is also used. Optimal hardware resources were used in the design of the hardware implementation, which aimed to minimize hardware complexity utilization. To top it all off, the design allows for great throughput by processing several channels simultaneously. Therefore, instead of processing individual heartbeats, arrhythmia categorization from compressed ECG data must be done in segments of predetermined length. Following this, we provide a deep learning (DL) model that, with the benefits of a high compression ratio (CR) and a low processing cost, can directly identify various forms of arrhythmia using compressed ECG segments of a set length. Our suggested strategy achieves an exact match rate of 97.03% at CR(Compression Ratio), according to experimental findings on the MITBIH arrhythmias database.