Short duration Vectorcardiogram based inferior myocardial infarction detection: class and subject-oriented approach

Area of Science:

• Biomedical Engineering
• Cardiology
• Signal Processing

Background:

• Myocardial infarction (MI) is a critical condition resulting from impaired blood circulation to the heart.
• Vectorcardiography (VCG) offers a 3-lead representation of the heart's electrical activity.
• Early and accurate detection of MI is crucial for patient outcomes.

Purpose of the Study:

• To develop an automated technique for detecting inferior myocardial infarction (MI) using short-duration VCG signals.
• To enhance the accuracy and efficiency of MI detection compared to existing methods.
• To validate the proposed method's reliability in real-world clinical scenarios.

Main Methods:

• Preprocessing of VCG signals using median and Savitzky-Golay filters.
• Time-invariant decomposition via Stationary Wavelet Transform (SWT).
• Feature extraction and selection using minimum-redundancy-maximum-relevance (mRMR), followed by supervised classification.

Main Results:

• Achieved high classification accuracy: 99.14% in a class-oriented approach and 89.37% in a subject-oriented approach.
• Demonstrated superior performance compared to current state-of-the-art methods.
• Utilized significantly shorter VCG signal segments for analysis.

Conclusions:

• The proposed VCG-based technique enables timely and reliable automated detection of MI.
• The method's high accuracy and efficiency, even with short signal durations, show significant clinical applicability.
• The subject-oriented approach confirms the technique's robustness and real-world reliability.