Objective Pain Assessment Using Wrist-based PPG Signals: A Respiratory Rate Based Method.

Area of Science:

• Biomedical Engineering
• Physiological Monitoring
• Pain Management

Background:

• Objective pain quantification is challenging in clinical settings due to dynamic and ambiguous pain phenomena.
• Respiratory rate, a bio-signal correlated with pain, is often overlooked due to monitoring difficulties.
• Existing pain assessment studies primarily focus on stimulated pain, not real clinical scenarios.

Purpose of the Study:

• To develop and validate an objective pain assessment method using respiratory rate from wristband-acquired Photoplethysmography (PPG) signals.
• To analyze PPG-derived respiratory rate in real post-operative patients.
• To establish a machine learning-based system for classifying pain levels.

Main Methods:

• Respiratory rate was derived from post-operative patients' PPG signals using Empirical Mode Decomposition (EMD).
• Statistical features were extracted from the respiratory rate, followed by feature selection.
• A weak supervision method was employed to handle imbalanced data labels.
• Machine learning algorithms performed binary classification (No Pain vs. Pain Levels 1-3).

Main Results:

• The proposed method achieved high prediction accuracies: up to 81.41% (NP vs. PL1), 80.36% (NP vs. PL2), and 79.48% (NP vs. PL3).
• Performance metrics surpassed existing state-of-the-art results.
• The study successfully utilized data from real post-operative patients.

Conclusions:

• Objective pain assessment is feasible using respiratory rate extracted from wristband PPG signals in post-operative patients.
• The developed method offers a non-invasive and accurate approach to pain monitoring in real clinical environments.
• This technique has the potential to significantly improve post-operative pain management and patient care.