Abstract
Myocardial infarction is an irreversible damage of heart muscle caused by prolonged oxygen deficiency. As a result, the presence of damaged tissue will alter the normal sinus rhythm. Hence, the paper proposes to profile history of myocardial infarction from electrocardiogram using artificial neural network. Data for anterior and inferior myocardial infarction, as well as healthy control is acquired from PTB Diagnostic ECG Database. Subsequently, QRS power ratio features for different frequency zones are extracted from the pre-processed electrocardiogram. Discriminative ability of the features is assessed using k-nearest neighbor. The best combination of features with 99.7% testing accuracy is the power ratio composite that combines both low-frequency and mid-frequency information. An intelligent profiling model is successfully developed using the composite features and an optimized artificial neural network. The model was able to identify between different electrocardiogram groups with overall accuracy of 98.4% and mean squared error of less than 0.1. Conclusively, the proposed signal processing approach has provided an improved alternative to the established methods from literature.
Highlights
Electrocardiogram (ECG) is non-invasive electrical recording of the heart
A more thorough investigation is proposed by focusing on different frequency zones of the QRS complex; lowfrequency (5–15 Hz), mid-frequency (15–80 Hz) and highfrequency (150–250 Hz) components [5]
By incorporating signal processing and intelligent modelling techniques, the study evaluates the feasibility of QRS power ratio features in different frequency zones for profiling ECG with history of myocardial infarction (MI)
Summary
Electrocardiogram (ECG) is non-invasive electrical recording of the heart. The bio-potential signal arises from propagation of ionic impulses throughout the cardiac conduction system. There has been an attempt to investigate the ECG of patients who survived acute MI This was based on the assumption that the damaged myocardium would introduce irregularities to the sinus rhythm. The study focused on the power ratio features from the bipolar and augmented limb leads. The method is advantageous as it is capable of learning and can generalize solutions to a given problem [9] Far, both kNN [10, 11] and ANN [12, 13] have been widely used to classify features and model complex non-linear relationships for various biomedical applications. By incorporating signal processing and intelligent modelling techniques, the study evaluates the feasibility of QRS power ratio features in different frequency zones for profiling ECG with history of MI. The preceding analyses defined the behavior of QRS complex within a single frequency zone (1– 20 Hz) [14]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: International Journal of Engineering & Technology
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
