Abstract
We have developed a new method for estimation of breathing frequency using spectrum analysis of fluctuation of R wave amplitude. In this article, we introduced the algorithm for estimation of breathing frequency and examined a reliability of that for five healthy subjects. In the first experiment, subjects controlled their respiration at six patterns of frequencies: 0.10, 0.17, 0.25, 0.33, 0.50Hz and breath holding as a reference value. Each of breathing frequencies was continued for 2 minutes. Electrocardiogram was measured continuously during examination. It was processed with personal computer after tests; we estimated breathing frequency with spectrum analysis of fluctuation of R wave amplitude. We examined reliability to be compared an estimated value with the reference value. In the results, although the estimated value was same as reference value for all subjects at 0.10 and 0.17Hz, a slight error was occurred in range from 0.01 to 0.02Hz at 0.25, 0.33 and 0.50Hz. In the second experiment, we examined the relation between tidal volume and the estimation error. In addition, we tested the effect of noise with activity of daily living. Eleven healthy subjects controlled their tidal volume at three grade: small, normal and large. Each of respiration patterns was continued for 2 minutes. They also walked on the treadmill at 4km/h for 2 minutes. We examined reliability to be compared the estimated value with the value of respiratory gas analyzer. In the results, the estimated value was almost same as reference value for normal and deep breathing. The relationship between tidal volume and the estimation error was not clear. During walking, the estimated breathing frequency was almost same as value of gas analyzer. In conclusion, it was proved that the estimation of breathing frequency with spectrum analysis of fluctuation of R wave amplitude had high reliability.
