A lossless multichannel bio-signal compression based on low-complexity joint coding scheme for portable medical...

Area of Science:

• Biomedical Engineering
• Signal Processing
• Data Compression

Background:

• Increasing demand for high-quality personal multichannel medical signal compression in real-time health systems.
• Existing MPEG-4 audio lossless coding (ALS) offers joint channel-coding for multichannel signals but suffers from high computational complexity.
• Need for efficient and low-complexity compression solutions for portable medical devices.

Purpose of the Study:

• To present a novel multichannel hardware encoder for medical signal compression.
• To reduce the computational complexity of multichannel signal encoding while maintaining or improving compression performance.
• To enable efficient data compression for portable personal medical product applications.

Main Methods:

• Development of a low-complexity joint-coding technique and a shared multiplier scheme for hardware implementation.
• Modification of the joint-coding decision method based on residual signal cross-correlation and compression ratio.
• Implementation of a reference channel selection scheme for optimized entropy coding.
• Hardware encoder designed to operate at 40 MHz with two-channel parallel encoding capability.

Main Results:

• The proposed hardware encoder achieved a 0.06% increase in compression ratio compared to the MPEG-4 ALS reference software.
• Computational complexity was reduced by 20.72% compared to the MPEG-4 ALS reference software.
• Compression ratio improved by approximately 11.92% compared to single-channel bio-signal lossless data compressors.

Conclusions:

• The developed low-complexity multichannel hardware encoder offers a viable solution for efficient medical signal compression in portable devices.
• The optimized joint-coding and reference channel selection strategies effectively reduce computational load without significant loss in compression performance.
• This technology supports the advancement of real-time health systems and personal medical monitoring applications.