Improving Respiratory Timing Estimation Using Quality Indexing and Electrocardiogram-Derived Respiration.

Area of Science:

• Physiology
• Biomedical Engineering

Background:

• Accurate respiratory timing is crucial for many applications.
• Respiratory effort (RSP) belts are common but susceptible to artifacts from looseness and movement.
• These artifacts can corrupt signals and lead to inaccurate respiration rate (RR), inspiration time (Ti), and expiration time (Te) estimations.

Purpose of the Study:

• To demonstrate how respiration quality indexing and outlier removal can mitigate RSP signal artifacts.
• To evaluate the impact of these methods on the accuracy of RR, Ti, and Te estimation.
• To assess the utility of ECG-derived respiration (EDR) in improving respiratory timing accuracy.

Main Methods:

• Collected RSP and ECG signals from 15 healthy participants during controlled breathing exercises in various postures.
• Processed RSP signals to estimate breath-by-breath RR, Ti, and Te.
• Compared RSP-derived estimates against spirometry using Bland-Altman analysis.
• Incorporated quality indexing, outlier removal, and EDR for artifact handling.

Main Results:

• Quality indexing and outlier removal improved 95% limits of agreement for respiratory timing by 10-40%.
• Using EDR during RSP signal artifacts significantly reduced the amount of data removal needed for accurate estimations.
• These enhancements were consistent across all tested postures.

Conclusions:

• Respiration quality indexing and outlier removal are effective in improving the accuracy of RSP-based respiratory timing.
• EDR provides a valuable complementary signal to enhance robustness when RSP signals are compromised.
• These techniques collectively improve the reliability of RR, Ti, and Te measurements, aiding respiratory health monitoring.