Energy-Based DCT Approach for PPG Compression.

Area of Science:

• Biomedical Engineering
• Signal Processing
• Wearable Technology

Background:

• Photoplethysmography (PPG) is vital for remote health monitoring via wearable devices.
• Resource constraints in wearables necessitate efficient PPG signal compression.
• Maintaining the clinical morphology of PPG signals during compression is crucial.

Purpose of the Study:

• To develop an efficient PPG compression technique.
• To integrate noise reduction within the compression algorithm.
• To optimize storage and power consumption for wearable health devices.

Main Methods:

• A novel PPG compression approach utilizing the energy of discrete cosine transform (DCT) was developed.
• The method was evaluated on multiple public databases, including MIMIC-II.
• Performance metrics included compression ratio and percentage root-mean-square difference (PRD).

Main Results:

• An average compression ratio of 37.46 was achieved on the MIMIC-II database.
• A low percentage root-mean-square difference of 0.0688 was recorded.
• The proposed method significantly outperformed existing PPG compression techniques.

Conclusions:

• The DCT energy-based approach offers superior compression and denoising for PPG signals.
• This algorithm optimizes memory usage and preserves clinically relevant PPG fiducial points.
• It supports enhanced ubiquitous health monitoring with faster diagnostics and feedback.