Justificación de las decisiones…

The development of several BCG systems embedded in weighing scales [66] in the last years has prompted a renewed interest on this signal because these devices are already common in the daily routine of many homes and are periodically used to monitor the health status from the body weight.

The BCG waveform is originated from the forces caused by cardiac ejection of blood hence any interval measured from the ECG QRS complex to the BCG depends on the PEP, the time from the electric activation of the heart to the beginning of cardiac ejection (see Fig. 1.3). As the physiological origin of the BCG waves is still controversial, the J wave is usually the reference BCG

78 Assessment of trends in the cardiovascular system

waveform feature used to assess the PEP because its SNR is higher than that of other BCG waves. The capability of the RJ interval, measured from the ECG R wave instead of the Q wave because of its higher SNR, in assessing the PEP [70] and contractility [71], which is inversely proportional to the PEP under stable conditions of preload and afterload, is good. The RJ can also be used to assess changes in systolic BP [14], which are strongly correlated with changes in PEP during short effort tests [72], and the strong correspondence between the RJ interval and systolic BP during the four phases of a Valsalva maneuver was reported in [C6].

Despite of the strong correlations reported, the performance of the RJ interval in assessing the PEP relies on the underlying hypothesis that the timing from cardiac ejection to the J wave is constant, which is not supported by any reported model or systematic observation and cannot be assumed to be true in all circumstances. BCG waves result from a superposition of forces caused by events during the propagation of the pressure-pulse along the arterial tree hence it is reasonable to assume that the later waves, which are originated by events in distal sites, are dependent on the PTT besides of the cardiac-ejection timing. Fig. 4.1 shows a representation of the ECG, BCG, and the BP at the aorta, carotid, and femoral arteries with different intervals of interest measured between its waves.

Fig. 4.1. Representation of the ECG, aortic BP, carotid BP, femoral BP, and the BCG. The RJ interval typically lasts about 210 ms [106] whereas the PEP usually lasts about 100 ms [18]. If it is assumed that the interval between the Q and R is roughly half of the QRS, which typically lasts about 80 ms [18], the J-wave timing appears to be about 150 ms later than the opening of the aortic valve, which is similar than the pressure-pulse timing at the finger or the femoral artery (see Fig. 3.8 and Fig. 3.20) and suggests that the wave is influenced by the PTT in addition to the PEP.

Earlier waves such as the H wave and the I wave should include a lesser amount of PTT and better assess the cardiac-ejection timing. However, the H wave, whose use to assess the PEP was

79 Time intervals from the BCG

suggested in [74] and is nearly simultaneous with the aortic valve opening, is better attributed to ventricular movements due to its absence in recordings without ventricular contraction [75]. Alternatively, the I-wave timing is typically about 90 ms after the ECG R wave [106], which is only about 30 ms after cardiac ejection and is similar to the pressure-pulse timing at the carotid artery (see Fig. 3.10), hence it is reasonable to assume that the I-wave timing is less influenced by superposed earlier waves and the PTT. Therefore, the I-wave timing should assess the cardiac- ejection timing better than the J-wave timing in spite of the higher SNR of this wave.

Further, the observations made above imply that the difference in timings of the I and J waves of the BCG, that is, the IJ interval, should be sensitive to the PTT in a manner analogous to other timing measurements from other proximal to distal signals. Therefore, as the I-wave timing and the J-wave timing are dependent to the carotid ejection timing and the J wave presumably includes a higher amount of PTT, it is reasonable to assume that the difference is sensitive to the PTT, and more particularly to changes in aortic PTT, as the greatest vertical forces recorded in the longitudinal BCG are originated there. The different dynamics of the RJ, RI, and IJ intervals during a paced respiration maneuver reported in [C7] suggest that the intervals could be sensitive to different cardiovascular parameters.

Finally, the ability of the BCG in assessing the cardiac-ejection timing has still not been fully exploited since the signal can be used to measure the PTT combined with a distal pressure-pulse waveform integrated in a weighing scale such as an IPG. Similar measurement have not been performed until very recent times [76], even though some systems reported are theoretically capable of performing this measurement [15] and use the information from both signals to improve heartbeat detection [123][124]. The results reported in section 3.4.2 suggest that the IPG from foot- to-foot is a very interesting signal that can be obtained from a weighing scale, as the pressure-pulse timing of this signal is insensitive to PTT changes in limbs and therefore the interval measured from the BCG I wave to it should be sensitive only to the aortic PTT, which is of great interest since this is currently the gold standard measurement of PTT for the assessment of regional arterial stiffness [17].