Studying Dynamical Characteristics of Oxygen Saturation Variability Signals Using Haar Wavelet

Area of Science:

• Biomedical Signal Processing
• Complexity Analysis
• Physiological Monitoring

Background:

• Biomedical signal analysis aims to understand physiological complexity for health monitoring and disease prediction.
• Entropy-based measures are common for quantifying biomedical signal complexity, but novel methods are needed.
• Oxygen Saturation Variability (OSV) signals offer insights into physiological states.

Purpose of the Study:

• To introduce a novel Haar wavelet-based technique for analyzing OSV signal complexity.
• To evaluate the technique's ability to differentiate OSV signals during COVID-19 infection and after recovery.
• To compare the proposed method against established scale-based entropy measures.

Main Methods:

• Utilized Haar wavelets to analyze the complexity of OSV signals from 44 COVID-19 patients.
• Collected OSV signal recordings during acute illness and post-recovery phases.
• Compared the novel Haar wavelet technique with Multiscale Entropy (MSE), Multiscale Permutation Entropy (MPE), Multiscale Fuzzy Entropy (MFE), and Multiscale Amplitude-Aware Permutation Entropy (MAMPE).

Main Results:

• The proposed Haar wavelet algorithm demonstrated superior accuracy in classifying OSV signals between COVID-19 illness and recovery states.
• The technique was also more time-efficient than the compared entropy measures (MSE, MPE, MFE, MAMPE).
• Preliminary results from this pilot study indicate significant potential for the Haar wavelet approach.

Conclusions:

• The novel Haar wavelet technique shows promise as an effective and efficient tool for biomedical signal complexity analysis.
• This method outperforms traditional entropy measures for differentiating COVID-19 infection states based on OSV signals.
• Further research is warranted to validate the algorithm on larger datasets and for other biomedical signal classification tasks.