Sabherwal Pooja, Agrawal Monika, Singh Latika
The NorthCap University, Gurgaon, India.
CARE, IIT Delhi, New Delhi, India.
Cardiovasc Eng Technol. 2023 Apr;14(2):167-181. doi: 10.1007/s13239-022-00643-1. Epub 2022 Sep 26.
In the ECG signals, T-waves play a very important role in the detection of cardiac arrest. During myocardial ischemia, the first significant change occurs on the T-wave. These waves are generated due to the repolarization of the heart ventricle. The independent detection of T-waves is a bit challenging due to its variable nature, therefore, most of the algorithms available in the literature for T-wave detection use the detection of the QRS complex as the starting point. But accurate detection of Twave is very much required, as clinically, the first indication of a shortage of blood supply to the heart muscle (myocardial ischemia) shows up as changes in T-wave followed by other changes in the morphology of the ECG signal.
In this paper, an efficient and novel algorithm based on Continuous Wavelet Transform (CWT) is presented to detect the Twave independently. In CWT, for better matching, a new mother wavelet is designed using the pattern and shape of the Twave. This algorithm is validated on all the signals of the QT database.
The algorithm attains an average sensitivity of 99.88% and positive predictivity of 99.81% for the signals annotated by the cardiologists in the database.
在心电图信号中，T波在心脏骤停检测中起着非常重要的作用。在心肌缺血期间，T波会发生第一个显著变化。这些波是由于心室复极化产生的。由于T波性质多变，独立检测T波具有一定挑战性，因此，文献中大多数用于T波检测的算法都以检测QRS复合波作为起点。但准确检测T波非常必要，因为在临床上，心肌供血不足（心肌缺血）的首个迹象表现为T波变化，随后是心电图信号形态的其他变化。
本文提出一种基于连续小波变换（CWT）的高效新颖算法，用于独立检测T波。在连续小波变换中，为了更好地匹配，根据T波的模式和形状设计了一种新的母小波。该算法在QT数据库的所有信号上进行了验证。
对于数据库中心脏病专家标注的信号，该算法的平均灵敏度达到99.88%，阳性预测值达到99.81%。
