Classification of pulse waveforms using edit distance with real penalty
EURASIP Journal on Advances in Signal Processing
| Hi-index | 0.00 |
Arterial stiffness is an important determinant of cardiovascular risk and, consequently, it is necessary to develop a system which measures the central aortic blood pressure noninvasively. Recently, the aortic augmentation index (AIx), which can be calculated from the estimated aortic waveform based on the radial pressure waveform, has been used as a quantitative index of systemic arterial stiffness. The pressure waveform at various body sites is a composite of the forward and reflected waves and, therefore, the relationship between the aortic pressure and peripheral pressure is not fixed. Consequently, it is necessary to establish an algorithm for obtaining the central aortic pressure waveforms from the radial pressure waveforms. Transfer functions were established using a 10th order ARX model by setting the radial pulse as the input and the aortic pulse as the output, and their stability was verified through residual analysis. The objective of this study was to compare the central blood pressure values estimated from the radial pressure waveform using the ARX model with the central pressure waveforms acquired invasively from patients undergoing angioplasty using a RADI pressure wire. Fifty subjects (35 males and 15 females) participated in the study. The central aortic pressure waveforms obtained using the RADI pressure wire system and the radial arterial pulse waves obtained using a pressure sensor were acquired simultaneously. The results show that there was a linear relationship between the systolic pressure measured by the RADI pressure wire and that estimated using the ARX transfer function (r = 0.848). Also, the pulse pressures measured by the RADI pressure wire were linearly related to those estimated using the ARX transfer function (r = 0.893). Keywords: central aortic pressure, arterial stiffness, augmentation index, ARX model