Koski A
Department of Computer Science, University of Turku, Finland.
Comput Methods Programs Biomed. 1997 Jan;52(1):23-33. doi: 10.1016/s0169-2607(96)01779-8.
We have studied the lossless encoding of ECG signals. With suitable code we aim to reduce the storage space needed by ECG signals. Several methods designed for ECG compression use lossy, i.e. irreversible techniques, in which the original signal is lost, but the restored approximation is almost equal to the original. We aimed, however, to use reversible methods which are able to restore the original signal exactly. We have examined various methods and developed a new approach based on structural recognition and extraction of ECG complexes. Comparative conclusions are drawn from the compression efficiency of lossless and lossy methods. In this study, the effect of sampling frequency, resolution and filtering is also examined.
我们研究了心电图（ECG）信号的无损编码。通过合适的编码方式，我们旨在减少ECG信号所需的存储空间。为ECG压缩设计的几种方法采用了有损（即不可逆）技术，在这些技术中，原始信号会丢失，但其恢复后的近似值与原始信号几乎相同。然而，我们的目标是使用能够精确恢复原始信号的可逆方法。我们研究了各种方法，并基于ECG复合波的结构识别和提取开发了一种新方法。从无损和有损方法的压缩效率得出了比较结论。在本研究中，还考察了采样频率、分辨率和滤波的影响。
