The work studies the feature extraction from wide band electrocardiogram (ECG) signals for classifying myocardial infarction (MI) stages.
The data for the analysis was taken from Physikalisch Technische Bundesanstalt diagnostic ECG database.
Multivariable autoregressive modeling technique was employed, and its coefficients were used as features.
The orthogonal ECG signals and standard ECG signals with different frequency ranges were generated for the analysis.
MI classification accuracy can be improved by introducing higher frequency components.