Automatic Detection of Seismocardiogram Sensor Misplacement for Robust Pre-Ejection Period Estimation in Unsupervised
1School of Electrical and Computer Engineering at the Georgia Institute of Technology, Atlanta, GA 30332 USA.
IEEE Sensors Journal
|November 1, 2017
Summary
Sensor placement affects seismocardiography (SCG) accuracy for heart failure monitoring. This study introduces a method to detect incorrect SCG sensor positioning, ensuring reliable pre-ejection period (PEP) estimation for home use.
Area of Science:
- Biomedical Engineering
- Cardiovascular Physiology
- Signal Processing
Background:
- Seismocardiography (SCG) measures chest vibrations to assess myocardial contractility via the pre-ejection period (PEP).
- SCG shows promise for home monitoring of heart failure patients by classifying decompensated from compensated states.
- Accurate PEP extraction relies on SCG signals correlated with reference methods like impedance cardiography (ICG), typically using "ideal" sensor placement.
Purpose of the Study:
- To investigate the impact of SCG sensor placement variations on the accuracy of pre-ejection period (PEP) estimation.
- To develop and validate a method for automatically detecting non-ideal SCG sensor positioning to ensure reliable systolic time interval estimation.
Main Methods:
- Analyzed how changes in SCG sensor position on the chest affect the regression curves used for PEP estimation.
- Developed a classification algorithm to distinguish between ideal and non-ideal SCG sensor placements.
- Tested the classifier using SCG data from rest and exercise recovery to differentiate positioning effects from physiological variations.
Main Results:
- Demonstrated that SCG sensor position significantly alters the regression curve for PEP estimation.
- The developed classification algorithm achieved 0.83 precision and 0.82 recall in detecting correct sensor placement.
- The method effectively discriminated between waveform changes due to sensor positioning and those due to physiological effects during rest and exercise recovery.
Conclusions:
- SCG sensor placement is critical for accurate PEP estimation and reliable heart failure monitoring.
- The proposed detection method can mitigate errors arising from suboptimal sensor positioning in unsupervised settings like home monitoring.
- This technique enhances the robustness and clinical utility of SCG for remote patient management.
