In pulse oximetry the principles of photoplethysmography are used for determination of heart rate and in some devices to also display the volume pulse. It has been suggested, that a more detailed analysis of the signal may allow quantitative analysis of peripheral hemodynamic events. We describe a new computer assisted time discrete analysis of the volume pulse, studying its reliability and the method s fundamental assumption of a linear relationship between changes in amplitude and changes in time sequences of the volume pulse.
Method: In a finger clip two diodes emit near infrared (840 nm, NIR) and red light (640nm, RED) into the finger tip, where it is remitted mainly by the erythrocytes. 70 sec of recorded signal is filtered and the resulting volume pulse analysed off-line using a computer. On each volume pulse the time of the first (Tmax), the second maximum (time of dicrote, Td) and the duration of the volume pulse (Tp) are measured and the mean values displayed. In addition, the fundamental arterial oscillation Tag = Td - Tmax and all the above values in relation to Tp are calculated. Using NIRP, 54 healthy young volunteers (19 female, mean age 27.0 +/- 3.4 years) were studied and the individual mean values calculated from 960 measurements. The reliability during 10 repetitive measurements was investigated in 26 of the 54 volunteers. In 12 subjects 5 repetitive measurements were obtained from each finger and compared with each other. In 11 subjects the linear relationship between amplitude and time sequence was tested on > 30 000 single volume pulses. The finger clip photoelectrode was levelled with the right atrium in all measurements, skin close to the clip and room temperature were recorded.
Results: From the mean individual values the following time discrete values were calculated for the NIR signal (n = 41): Tp = 882.3 +/- 142.6 ms, Tmax = 214.8 +/- 28.3 ms, Td = 452.7 +/- 32.4 ms, Tdec = 667.4 +/- 133.6 ms, Tag = 237.9 +/- 36.3 ms, Tmax/Tdec = 0.34 +/- 0. 07, Td/Tdec = 0.7 +/- 0.11. For each parameter the individual standard deviation during 10 repetitive measurements (26 subjects) ranged between 2.2 and 6.1%. The time sequences found were not significantly different between the individual fingers. A linear relationship between changes in time sequence and changes in amplitude was found in all tested subjects (mean r = 0.96).
Conclusion: These results show, that the values obtained with time discrete NIRP are highly reproducible and show an individual SD of less than 6.5% under steady state conditions. The linear relationship between time sequence and amplitude found in the present study has to be confirmed in further studies on patients with pathologies of the macro- and microcirculation.