Development of an automatic Doppler flow signal detection system: Variability of pulmonary and aortic peak flow velocity

Purpose. Automatic Doppler flow signal detection systems can provide beat-to-beat information for large blood vessels. We have developed new equipment for automatic measurement of Doppler flow signals. The reliability of the system was examined, and the variability of aortic and pulmonary peak flow velocity was determined. Methods. We measured peak flow velocity using a newly developed system in healthy volunteers and patients with atrial fibrillation. Analysis of variability of peak flow velocity was performed with maximal entropy methods. Results. In Bland-Altman plots, the mean and standard deviation (SD) of differences in aortic peak flow velocities between the automatic and manual measurements were 0.22 ± 0.75cm/s and 0.85 ± 0.38cm/s, respectively, in five normal volunteers. Moreover, less than 5% of the plotted points were beyond ± 2 SD of the differences. Furthermore, good reproducibility was demonstrated using Bland-Altman plots and Pearson's correlation analysis. Identical reliability was obtained in patients with atrial fibrillation. The same results were obtained for pulmonary peak flow velocity. In five healthy subjects, aortic and pulmonary peak flow showed standard deviations of 7.2 ± 2.4 and 3.8 ± 0.6cm/s, respectively, and coefficients of variation of 6.1% ± 1.0% and 5.1% ± 1.1%, respectively, in time-domain variability. Similarly, frequency-domain variability was obtained for both peak flow velocities. Conclusion. The present study demonstrated the reliability of a newly developed automatic Doppler flow signal detection system. Using this system, the present study demonstrated for the first time aortic and pulmonary peak flow velocity variability. The present analytical methods may have considerable potential for studying aortic and/or pulmonary flow variability in connection with cardiac performance and prognosis of cardiac disease.