Algorithm for Quantifying Frontal EMG Responsiveness for Sedation Monitoring

Area of Science:

• Biomedical Engineering
• Intensive Care Medicine
• Neurophysiology

Background:

• Facial electromyographic (FEMG) activity is a potential indicator of patient response in intensive care units (ICUs).
• Quantifying FEMG activity requires robust algorithms for reliable clinical application.
• Monitoring sedation levels in ICU patients is crucial for optimizing care and patient outcomes.

Purpose of the Study:

• To investigate stimulation-related FEMG activity in ICU patients.
• To develop and optimize an algorithm for quantifying FEMG activity.
• To assess the algorithm's utility in monitoring patient sedation states.

Main Methods:

• Studied FEMG response patterns to vocal stimulation in 17 ICU patients.
• Collected continuous FEMG data from 30 ICU patients.
• Developed the Responsiveness Index (RI) algorithm and compared its values with clinical sedation assessments.

Main Results:

• Patients responding to vocal stimuli showed a significant increase in poststimulus FEMG power.
• Nonresponding patients exhibited no significant change in FEMG power.
• The RI algorithm demonstrated 0.90 sensitivity and 0.79 specificity for detecting deep sedation.

Conclusions:

• Consistent FEMG patterns associated with stimulation were identified in ICU patients.
• A simple and robust RI algorithm was successfully developed.
• The algorithm achieved good correlation with clinical sedation scores in the development dataset.