Robust Automatic Detection of P Wave and T Wave in Electrocardiogram

Abstract

The Electrocardiogram (ECG) is a significant tool to investigate the electrical activity of the heart. Even though P and T waves reveal very useful information, their accurate detection is a difficult and challenging task on account of noise, baseline drift and odd morphologies.The need for a rapid and effective detection algorithm motivated us to propose two automated methods: the percentile based Automatic Detection (pAD) and the Graphical based Automatic Detection (gAD). Both algorithms use statistical and probabilistic concepts to achieve adequate delineation and detection of the waves. The former uses the percentile as an adaptive threshold to define the location of these waves. The latter uses a "feature wavebank" to train a graphical probabilistic model named as Hidden Conditional Random Field (HCRF). The gAD algorithm takes advantage of the monotonicity and the slope of an ECG to detect and collect waves, which imports to the graphical model and classifies them to P or T waves.The efficiency of our proposed algorithms has been evaluated on 10 long-term (24-hour) ECG recordings of MIT-BIH Normal Sinus Rhythm Database. The training set we used for gAD was very small, only the 0.2% of the total number of the available waves. The results show a significant and promising detection accuracy rate. IntroductionThe Electrocardiogram represents a non-invasive method which provides crucial information about the functionality and performance of human heart and the influencing factors. The ECG waveform is well characterized by its cyclic behavior of sequential occurrence of P (atrial depolarization), QRS (ventricular depolarization), and T (ventricular repolarization) waves. Obviously, each wave represents a specific physiological performance of the heart system. As a result, a cumulative analysis of ECG wave components is significant in diagnosis and prognosis of abnormalities (morphological pattern changes) or disorder of heart rate and requires the detection of these waves.In literature, there is a growing number of researches on automatic analysis of ECG signal and delineation of P and T waves. The detection task is challenging due to amplitude and morphology variability, low signal-to-noise ratio (SNR), low amplitudes, and possible overlapping of the P wave and T wave with the QRS complex. There have been different approaches on detection of P and T waves. However, building a reliable and powerful method for automatic ECG delineation is still far from being solved and remains a topic of main interest. 2.Data and Methods 2.1. DataThe efficiency of our proposed algorithms has been evaluated on 10 long-term (24-hour) ECG recordings of MIT-BIH Normal Sinus Rhythm Database. The subjects had no significant arrhythmia with age range from 20 to 50. The database is sampled at 128 Hz and the data is available at uniform intervals of 7.8125 msec [19].

Non-Standard Electrode Placement Strategies for ECG Signal Acquisition

“…QT interval is an important variable in following the status of the patient on potentially pro-arrhythmic drugs [ 40 ]. Still, the T wave itself proposes a challenge, which has been proposed to be solved, among others, through the application of wavelet transform [ 41 ], artificial neural networks [ 42 ], signal morphology [ 43 ], ensemble averaging [ 44 ], and derivatives of the signal [ 45 ].…”