Abstract
Electrical activity in the heart arises from the movement of positively and negatively charged ions across the cell membrane of cardiac cells. The movement of these ions occurs in an organized and repetitive fashion that generates a sequences of voltage changes called the action potential. The action potential activates the heart and initiates contraction. It also generates potential differences between different parts of the heart that can be recorded from the surface of the heart. It can also be recorded by electrodes attached to the surface of the skin, generating the electrocardiogram (ECG). Recording the ECG from multiple standardized positions allows clinically useful conclusions to be made regarding the sequence of activation of the heart and about the presence of pathology. Analog recording of ECGs has largely been supplanted by digital recording. The ECG signal is processed by high pass and low pass filters to allow maximal fidelity while minimizing electrical noise. A technique called common mode rejection rejects signals that appear at the same amplitude at each electrode since such signals most likely represent artifact. Analog to digital conversion typically occurs at a high sampling rate of 500 Hz. Some devices use variable sampling rates, reducing it for portions lower frequency portions of the signal. This preserves the fidelity of the ECG while creating smaller files. Many ECG recorders provide automated analysis of the waveform. To do this, they form a template of the ECG waveform using techniques such as signal averaging to create an average or median waveform, selecting fiduciary points to align the elements of the ECG waveform, and using features such as changes in amplitude and slope to identify key elements of the waveform. Complex decision trees are used by programs to make diagnostic statements regarding the waveform. While reliable, computerized ECG analysis programs have been outperformed by expert human readers. Since the greatest accuracy may be achieved by combining computerized interpretation with over-reading by humans, current standards call for physician review of all ECGs.
Original language | English (US) |
---|---|
Title of host publication | Practical Signal and Image Processing in Clinical Cardiology |
Publisher | Springer London |
Pages | 113-131 |
Number of pages | 19 |
ISBN (Print) | 9781848825147 |
DOIs | |
State | Published - Dec 1 2010 |
ASJC Scopus subject areas
- General Medicine