Automatic Classification of Anomalous ECG Heartbeats from Samples Acquired by Compressed Sensing

Area of Science:

• Biomedical Engineering
• Signal Processing
• Cardiology

Background:

• Electrocardiogram (ECG) signals are crucial for diagnosing heart conditions.
• Compressed sensing enables efficient ECG data acquisition.
• Accurate classification of anomalous heartbeats is vital for patient care.

Purpose of the Study:

• To propose a novel method for classifying anomalous heartbeats from compressed ECG signals.
• To leverage compressed sensing for efficient cardiac signal analysis.
• To improve the accuracy and efficiency of arrhythmia detection.

Main Methods:

• Feature extraction using Discrete Cosine Transform (DCT) coefficients of compressed ECG signals.
• Classification employing k-nearest neighbor (KNN) ensemble classifiers.
• Utilizing signals acquired via compressed sensing technology.

Main Results:

• Achieved a classification accuracy of 99.40% for anomalous heartbeats.
• Successfully classified five distinct classes of anomalous heartbeats.
• Demonstrated the effectiveness of DCT and KNN ensembles on compressed signals.

Conclusions:

• The proposed method offers a highly accurate approach for anomalous heartbeat classification.
• Compressed sensing combined with DCT and KNN ensembles is effective for ECG analysis.
• This technique holds potential for real-time cardiac monitoring and diagnosis.