Electrocardiogram (ECG) Signal Diagnosis Based on Component Extraction
Auf Abdul-Rahmaan Hasso

Auf Abdul-Rahmaan Hasso, B.SC, Department of Electrical and Electronics Engineering, Mosul Iraq.
Manuscript received on 12 October 2013 | Revised Manuscript received on 20 October 2013 | Manuscript Published on 30 October 2013 | PP: 97-104 | Volume-3 Issue-5, October 2013 | Retrieval Number: D1217093413/13©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: This work presents a diagnosis system of ECG signal based on its component extraction. The ECG signal was analyzed in time & frequency domain techniques. In time domain techniques, the signal is segmented to extract all the medically important features that were used in the diagnosis. A bottom-up derivative-based algorithm was adopted. This Algorithm subjects the signal derivative to some empirical thresholds. The result of this method is a segment locating waveform that separates and delimits the various segments of the ECG. In frequency domain techniques, the signal is transformed by Fast Fourier Transformation. The signal is transformed sometimes beat by beat. The signal is analyzed in frequency domain by study the power spectrum and find thresholds for normal cases then compare these thresholds with other ECG signals to recognize the abnormal cases. Each disease has its own power spectrum which is different from the normal cases by a threshold in a specify location in the spectrum. Different medical criteria of diseases categories were used in making the diagnostic decision. They were taken from medical books. The system was tried on a large number of ECG signals, some samples of results were given as diagnostic reports.
Keywords: ECG, Time Domain, FFT, Diagnosis.

Scope of the Article: Component-Based Software Engineering