Analyzing Seismocardiogram Cycles to Identify the Respiratory Phases

Area of Science:

• Biomedical Engineering
• Cardiovascular Physiology
• Signal Processing

Background:

• Seismocardiogram (SCG) signals capture cardiac vibrations via sternal accelerometry.
• Identifying respiratory phases within SCG cycles is crucial for detailed cardiac analysis.

Purpose of the Study:

• To develop and validate a method for accurately identifying respiratory phases (inhale/exhale) in SCG signals.
• To assess the method's performance using both leave-one-subject-out and subject-specific scenarios.

Main Methods:

• SCG signals from 19 healthy subjects were collected and preprocessed.
• Time and frequency domain features were extracted by dividing SCG cycles into equal-sized bins.
• Support Vector Machines were used for feature selection and respiratory phase identification.

Main Results:

• Time-domain features demonstrated superior performance in identifying respiratory phases.
• Key features included characteristic points related to aortic valve events and cardiac cycle length.
• Average identification accuracies reached 88.1% (LOSO) and 95.4% (SS).

Conclusions:

• The developed method is efficient, reliable, and accurate for identifying respiratory phases in SCG cycles.
• This technique can improve the extraction of clinically significant data, such as systolic time intervals.