Time-domain ECG signal analysis based on smart-phone

Area of Science:

• Biomedical Engineering
• Signal Processing
• Mobile Health

Background:

• Electrocardiogram (ECG) analysis is crucial for diagnosing cardiac conditions.
• Real-time, accessible ECG monitoring on mobile platforms is increasingly important.
• Distinguishing between ventricular tachycardia (VT) and ventricular fibrillation (VF) is critical for patient outcomes.

Purpose of the Study:

• To develop and implement a time-domain ECG analysis algorithm on a smartphone.
• To accurately detect and classify normal heartbeats and premature ventricular contractions (PVCs).
• To efficiently differentiate between ventricular tachycardia (VT) and ventricular fibrillation (VF) using novel methods.

Main Methods:

• Utilized the Pan-Tompkins algorithm for QRS detection and beat classification.
• Implemented a computationally efficient method combining Lempel and Ziv complexity analysis with K-means for VT/VF separation.
• Developed a new classification rule for recognizing VT and VF.

Main Results:

• The system successfully detected and classified normal beats and PVCs.
• The proposed method demonstrated efficient separation of VT and VF.
• The algorithm architecture showed good performance on the MIT-BIH Database.

Conclusions:

• A novel, efficient time-domain ECG analysis algorithm suitable for mobile platforms has been developed.
• The system offers accurate detection and classification of various arrhythmias.
• This mobile-based approach has significant potential for remote cardiac monitoring and diagnosis.