In this approach, as the frequency ( f) increases, the spectral bandwidth (6 σ f) increases. In a classic wavelet analysis, C is a constant (e.g., 7), ensuring an equal number of cycles in the mother wavelet for each frequency. The Morlet wavelet has a Gaussian shape that is defined by a ratio ( σ f = f/ C) and a wavelet duration (6 σ t), where f is the center frequency and σ t = 1/(2 πσ f).
#Morlet wavelet matlab code software#
Time-frequency analysis of the EEG gamma activity was based on Morlet wavelets using the freely distributed FieldTrip ( ) software in Matlab ( ), as in our prior work ( Roach and Mathalon, 2008). Mathalon, in Supplements to Clinical Neurophysiology, 2013 10.5.6 EEG time-frequency analysis
When there is still doubt about the normality assumption, one can apply a log or square-root transform ( Kiebel et al., 2005). These averaging operations render the contrasts or summary statistics normally distributed, by central limit theorem. However, in most cases, the contrasts have a near-normal distribution because of averaging over time, frequency and trials and, more importantly, taking differences between peristimulus times or trial-types. In general, we assume that the second-level error for contrasts of power is normally distributed, whereas power data per se have a χ 2-distribution. These contrasts can be modelled using repeated-measures ANOVA, where time and frequency are both factors with multiple levels.
#Morlet wavelet matlab code trial#
Alternatively, one can compute, per trial type and subject, several averages in the time-frequency plane and take them up to the second level. This might be an average in the time-frequency plane (e.g. When time and frequency are considered experimental factors, we compute contrasts at the first level and pass them to the second. Induced activity is computed by averaging P ij over trials and subtracting the power of the evoked response. Where z(t,ω)* ij is the complex conjugate.
0 kommentar(er)
