A sliding-window based algorithm to determine the presence of chest compressions from acceleration data

Area of Science:

• Emergency Medicine
• Biomedical Engineering
• Data Science

Background:

• Resuscitation quality assessment is crucial for improving patient outcomes.
• Manual annotation of chest compression data is time-consuming and prone to error.
• Automated analysis of resuscitation data can enhance research and clinical practice.

Purpose of the Study:

• To present an algorithm for automatic detection of chest compression periods.
• To validate the algorithm using detailed case data and expert annotations.
• To enable efficient analysis of large resuscitation datasets.

Main Methods:

• Utilized defibrillator records (accelerometry, ECG, capnography) from the German Resuscitation Registry.
• Involved expert physician consensus annotation of cardiac arrest, return of spontaneous circulation, and chest compression periods.
• Developed and applied an automated algorithm for chest compression detection.

Main Results:

• Successfully developed an algorithm that reliably detects chest compression periods automatically.
• The algorithm eliminates the need for laborious manual annotation.
• Demonstrated the algorithm's utility in analyzing complex resuscitation data.

Conclusions:

• The developed algorithm offers a significant advancement for automated resuscitation quality assessment.
• This tool facilitates machine learning approaches and the handling of big data in resuscitation research.
• The automated detection of chest compressions enhances the efficiency and scalability of resuscitation data analysis.