Accurate and consistent automatic seismocardiogram annotation without concurrent ECG.

Area of Science:

• Biomedical Engineering
• Cardiovascular Physiology
• Signal Processing

Background:

• Seismocardiography (SCG) measures heart vibrations, but precise cardiac timing relies on accurate fiducial point identification.
• Current SCG annotation requires concurrent electrocardiography (ECG) measurements, limiting its standalone use.
• Developing an algorithm for SCG annotation independent of ECG is crucial for advancing non-invasive cardiac monitoring.

Purpose of the Study:

• To design and validate an algorithm for automatic SCG annotation without concurrent ECG.
• To assess the accuracy of the algorithm in identifying cardiac mechanical timings under varying physiological conditions.
• To evaluate the potential of SCG as a standalone cardiovascular monitoring device.

Main Methods:

• 18 participants underwent graded lower body negative pressure (0 to -50 mmHg).
• Simultaneous ECG and SCG data were collected; R peaks from ECG were used for manual SCG annotation.
• An automated algorithm was developed to annotate SCG signals, and its performance was compared against manual annotations.

Main Results:

• The algorithm demonstrated high accuracy in annotating cardiac cycles across different negative pressure levels.
• Comparisons of isovolumetric moment timings between manual and algorithmic annotations showed high consistency.
• Analysis of LF/HF ratios derived from SCG-derived timings closely matched those from ECG R-peaks, indicating reliable autonomic function assessment.

Conclusions:

• The developed algorithm enables accurate SCG annotation without ECG, establishing SCG as a viable standalone heart monitoring tool.
• The algorithm's accuracy and consistency support its use in healthy individuals at rest.
• This technology holds promise for future applications in monitoring pathological cardiovascular conditions.