A joint application of optimal threshold based discrete cosine transform and ASCII encoding for ECG data compression...

Area of Science:

• Biomedical Engineering
• Signal Processing
• Data Compression

Background:

• Electrocardiogram (ECG) data requires efficient compression for storage and transmission.
• Existing methods often face trade-offs between compression ratio and signal fidelity.
• Data security during transmission is crucial for sensitive medical information.

Purpose of the Study:

• To develop a joint data compression and encryption technique for ECG signals.
• To leverage the energy compaction property of DCT and DPCM for improved performance.
• To validate the proposed method against the MIT-BIH database and compare with state-of-the-art approaches.

Main Methods:

• Utilizing Discrete Cosine Transform (DCT) for feature extraction and energy compaction.
• Applying Differential Pulse Code Modulation (DPCM) based quantization to DCT coefficients.
• Optimally thresholding coefficients using the Regula-Falsi method.
• Encoding the quantized sequence with Huffman coding and converting to ASCII for transmission.
• Implementing inherent encryption through the coding process.

Main Results:

• Achieved an average compression ratio (CR) of 18.03.
• Maintained a low average percent root mean square difference (PRD) of 1.06 and a high average quality score (QS) of 17.57.
• Demonstrated strong encryption capabilities with average entropy of 7.9692, Entropy Score (ES) of 0.9962, and correlation coefficient (rxy) of 0.0113.
• Outperformed contemporary methods in key performance metrics.

Conclusions:

• The joint DCT and DPCM approach offers a significant advancement in ECG data compression and encryption.
• The method provides a favorable balance between compression efficiency, signal fidelity, and data security.
• The proposed technique is well-justified by its superior performance metrics compared to existing methods.