Ensemble Empirical Mode Decomposition for atherosclerosis in high-risk subjects

Photoplethysmography (PPG) is a popular non-invasive method of waveform contour analysis in assessing arterial stiffness. Although stiffness index (SI) is a good parameter in assessing atherosclerosis, it dose not perform in aged and arterial stiffness patients, because of their ill-defined diastolic peaks in digital volume pulse (DVP) waveforms. To overcome the limitation, we use the mathematical model Ensemble Empirical Mode Decomposition (EEMD) to processing the record signals, the signals were decomposed into eight intrinsic mode functions (IMF 1-8), of which IMF5 was found to be the desired signal with a discernible diastolic peak. This study demonstrate a good agreement of SI_Original (calculate by original data) and SI_EEMD (calculated by IMF5). The results showed significant differences in SI_EEMD from the healthy aged and hypertension patients. Therefore, we successfully applied a mathematical model EEMD in processing the DVP signals, and assessed arterial stiffness in patients by SI_EEMD.