Dual-Sensor Signals Based Exact Gaussian Process-Assisted Hybrid Feature Extraction and Weighted Feature Fusion for...

Area of Science:

• Biomedical Engineering
• Signal Processing
• Physiological Monitoring

Background:

• Accurate respiratory rate (RR) estimation is crucial for patient and elderly care.
• Existing methods may lack reliability and uncertainty quantification.

Purpose of the Study:

• To develop a novel method for reliable RR estimation and uncertainty quantification.
• To improve the accuracy and stability of RR monitoring using physiological signals.

Main Methods:

• Utilized exact Gaussian process regression (EGPR) with hybrid feature extraction and fusion.
• Employed power spectral features and a multi-phase feature model for data compensation.
• Applied robust neighbor component analysis for feature selection and weighted fusion.

Main Results:

• The proposed EGPR method demonstrated improved reliability in RR estimation.
• Achieved the lowest mean absolute error (MAE) with EGPR-MF (0.993 bpm) and EGPR-feature fusion (1.064 bpm).
• The EGPR algorithm provides stable variation and confidence intervals for uncertainty estimation.

Conclusions:

• The EGPR-assisted hybrid feature extraction and fusion method offers a reliable approach for accurate RR estimation.
• This methodology is unique in providing stable variation and confidence intervals for RR monitoring.
• The proposed approach enhances patient monitoring through accurate and uncertainty-aware RR assessment.