This paper was intended to examine whether the power spectral density of the body sway in upright standing was always stable or not.
We used 8 healthy subjects, who were asked to stand on a force-plate with their eyes open or closed. One trial length was 3 minutes, and 5 trials carried out had intermissions of 5 minutes between each trial. Each trial was stored in a magnetic tape and fed into a computer in the form of digital values through an A/D converter. The sampling interval was 60 msec, so that the datum of each trial was transformed to 3, 072 digital values. One block was constructed 1, 024 digital values. Eight blocks extracted from one trial of 3, 072 points were obtained by moving their start adresses in 256 points steps, from 1 to 1, 793. Each block of 1, 024 points was transformed to the power spectrum by means of an F F T method, which was imposed 4 times of the Hanning window for smoothing. In this way, 8 power spectra were obtained from one trial and were named P
1 to P
8 in order of adress. The density of the power spectrum was expressed by logarithmic values.
A standard deviation calculated from 10 power spectra (P
3 and P
7 in each trial) of each subject was less than log
104 in all the frequency ranges computed, and also was not affected by two visual conditions, viz. eye open or closed. In order to investigate the time dependent variety, 8 power spectra from P
1 to P
8 were compared with the mean of 5 power spectra (P
3 in each trial) in all the frequency ranges They were named power-ratio-spectrum. The power-ratio-spectra were found to have values between 1/4 and 4 in either case of the anterior-posterior or the right-left sway.
Our results suggested on account of our analysis time and data length that there were no large varia-tions in the successive analysis
抄録全体を表示