Skordalakis E
Comput Biomed Res. 1984 Jun;17(3):208-21. doi: 10.1016/s0010-4809(84)80012-9.
A new method for recognizing noisy peaks in ECG waveforms is presented. This method is based on the principle that most noisy peaks in ECG waveforms appear as pairs of adjacent peaks which satisfy certain criteria. These pairs are called "noisy peak pairs" in this paper. The above principle was originally based on observation and was subsequently checked and found to be correct utilizing a standard ECG library, namely, the CSE library. With this method and using that library, 84.8% on the average of all peaks in every lead were recognized as noisy. This method is useful when employed as a step in the process of recognizing the real peaks in ECG waveforms.
提出了一种识别心电图波形中噪声峰值的新方法。该方法基于这样一个原理：心电图波形中的大多数噪声峰值表现为满足特定标准的相邻峰值对。本文中将这些对称为“噪声峰值对”。上述原理最初基于观察，随后利用一个标准心电图库，即CSE库进行了检验并发现是正确的。使用该方法并借助那个库，平均每条导联中所有峰值的84.8%被识别为噪声。当该方法用作识别心电图波形中真实峰值过程中的一个步骤时很有用。
