Hybrid method based on singular value decomposition and embedded zero tree wavelet technique for ECG signal...

Area of Science:

• Biomedical Engineering
• Signal Processing
• Data Compression

Background:

• Biomedical data, particularly ECG signals, requires significant storage and bandwidth.
• Real-time ambulatory and telemedicine systems face limitations due to large data quantities.

Purpose of the Study:

• To develop an efficient and simple technique for compressing ECG signals.
• To address the challenge of reducing data volume in ambulatory and telemedicine systems.

Main Methods:

• An algorithm combining Singular Value Decomposition (SVD) for initial low-rank matrix compression and Embedded Zero Tree Wavelet (EZW) for final compression.
• Utilized three beat segmentation approaches for 2-D ECG data array construction.
• Tested on the MIT-BIH arrhythmia database.

Main Results:

• Achieved a compression ratio of 24.25:1 for ECG signal Rec. 100.
• Maintained excellent signal reconstruction quality with a Percentage-Root-Mean-Square Difference (PRD) of 1.89%.
• Reduced data consumption from 3960bps to 162bps.

Conclusions:

• The proposed SVD and EZW based method is efficient and flexible for various ECG signals.
• The algorithm effectively controls reconstruction quality.
• Significant potential to reduce memory space in health data centers and save bandwidth in telemedicine.