Comparison of the fast Fourier transform and continuous wavelet transform for examining mechanomyographic frequency versus eccentric torque relationships

Abstract

The purpose of this study was to compare the eccentric torque-related patterns for mechanomyographic (MMG) center frequencies (mean power frequency (MPF), median frequency (MDF), and average instantaneous mean power frequency (AIMPF)) determined by the fast Fourier transform (FFT) and continuous wavelet transform (CWT). Eight adults (mean ± S.D. age = 22.5 ± 2.4 years) performed submaximal to maximal, eccentric isokinetic muscle actions of the biceps brachii on a Cybex 6000 dynamometer. The mean MMG MPF, MDF, and AIMPF values for both the absolute and normalized data from 10 to 100% eccentric peak torque (PT) were highly intercorrelated at r = 0.908–0.985. Linear models provided the best fit for the absolute MMG MPF ( r = 0.873), MDF ( r = 0.831), and AIMPF ( r = 0.924), as well as normalized MMG MPF ( r = 0.869), MDF ( r = 0.816), and AIMPF ( r = 0.920) versus percentage eccentric PT relationships. There were no significant differences ( p > 0.05) among the linear slope coefficients for the MMG MPF, MDF, and AIMPF versus percentage eccentric PT relationships for either the absolute or normalized data. These results suggested that Fourier or wavelet transform procedures can be used to examine the patterns of MMG responses during eccentric muscle actions of the biceps brachii.