A Novel Approach for Different Morphological Characterization of ECG Signal

Abstract

The earlier detection of Cardiac arrhythmia of ECG waves is important to prevent cardiac disorders. A good system depends heavily upon the precise and consistent estimate of ECG signal morphology i.e. QRS complex, T and P wave. From the benchmark data bases: MIT-BIH Arrhythmia, QT and European ST-T database the ECG is fetched then the noise is removed from the digitized ECG signal. To analyze various power spectrum of ECG signal Stationary Wavelet Transform (SWT) is applied to the de-noised signal. Based upon the spectrum QRS complex T and P waves are detected and also delineated using different amplitude threshold values. This gives simple and reliable method for the detection and delineation of the constituent waves from a given ECG signal has been the fundamental goal of automatic cardiac arrhythmia detection. This algorithm allows delineation of different morphologies of QRS complex P and T wave.

APPENDIX

The amplitude thresholds in the presented algorithms can be grouped into three types. First, the thresholds used to decide if a pair of maximum moduli with opposite sign can account for a wave: $\mathop \in \nolimits_{\text{QRS}}^{ 1}$,$\mathop \in \nolimits_{\text{QRS}}^{ 2}$, $\mathop \in \nolimits_{\text{QRS}}^{ 3}$, $\mathop \in \nolimits_{\text{QRS}}^{ 4}$ (for QRS detection), $\mathop \in \nolimits_{\text{T}}$ and $\mathop \in \nolimits_{\text{P}}$. (in the T/P wave delineation). These thresholds are proportional to the RMS value of the WT at the corresponding scales. For QRS detection, in each expert of $2^{16}$ samples.

$\in_{\text{QRS}}^{ 1}$ = RMS (${\text{W}}_{{ 2^{\text{i}} }} {\text{x[n]}}$), i = 1, 2, 3

$\in_{\text{QRS}}^{ 4}$ = 0.5RMS (${\text{W}}_{{ 2^{ 4} }} {\text{x[n]}}$).

For T and P waves

$\in_{\text{T}}$ = 0.25 RMS (${\text{W}}_{{ 2^{ 4} }} {\text{x[n]}}$) where the RMS is measured in each interval between

$\in_{\text{P}}$ = 0.02 RMS (${\text{W}}_{{ 2^{ 4} }} {\text{x[n]}}$) two consecutive QRS. The morphology of QRS complexes and the type of T/P wave depend of number of significant maximum moduli. The thresholds to determine if they are significant, ${{\upgamma}}_{{{\text{QRS}}_{\text{pre}} }}$, ${{\upgamma}}_{{{\text{QRS}}_{\text{post}} }}$, ${{\upgamma}}_{\text{T}}$ and ${{\upgamma}}_{\text{P}}$ are related to the amplitude of the global maximum modulus within the corresponding search window (sw).

${{\upgamma}}_{{{\text{QRS}}_{\text{pre}} }}$= 0.06 max($\left| {{\text{W}}_{{ 2^{ 2} }} {\text{x[n]}}} \right|$), n $\in {\text{SW}}_{\text{QRS}}$ A third group of thresholds are used to ${{\upgamma}}_{{{\text{QRS}}_{\text{post}} }}$= 0.09 max($\left| {{\text{W}}_{{ 2^{ 2} }} {\text{x[n]}}} \right|$), n $\in {\text{SW}}_{{_{\text{QRS}} }}^{{}}$ determine the onset/offset of QRS complex. They are proportional to the amplitude of the WT at the first/last maximum modulus of the complex or wave.

$$ \mathop {{\upxi}}\nolimits_{\text{QRSon}} = \left\{ \begin{gathered} 0. 0 5 {\text{W}}_{{ 2^{ 2} }} {\text{x[n}}_{\text{first}} ] , {\text{if\,W}}_{{ 2^{ 2} }} {\text{x[n}}_{\text{first}} ]> 0\hfill \\ 0. 0 7 {\text{W}}_{{ 2^{ 2} }} {\text{x[n}}_{\text{first}} ] , {\text{if\,W}}_{{ 2^{ 2} }} {\text{x[n}}_{\text{first}} ]< 0\hfill \\ \end{gathered} \right. $$

$$ \mathop {{\upxi}}\nolimits_{\text{QRSend}} = \left\{ \begin{gathered} 0. 1 2 5 {\text{W}}_{{ 2^{ 2} }} {\text{x[n}}_{\text{last}} ] , {\text{if\,W}}_{{ 2^{ 2} }} {\text{x[n}}_{\text{last}} ]> 0\hfill \\ 0. 7 1 {\text{W}}_{{ 2^{ 2} }} {\text{x[n}}_{\text{last}} ] , {\text{if\,W}}_{{ 2^{ 2} }} {\text{x[n}}_{\text{last}} ]< 0\hfill \\ \end{gathered} \right. $$