hmrR_MotionCorrectWavelet - BUNPC/Homer3 GitHub Wiki

UI Name :
Wavelet_Motion_Correction

Syntax :
data_dod = hmrR_MotionCorrectWavelet(data_dod, mlAct, iqr, turnon)

Description :
Perform a wavelet transformation of the dod data and computes the distribution of the wavelet coefficients. It sets the coefficient exceeding iqr times the interquartile range to zero because these are probably due to motion artifacts. set iqr<0 to skip this function. The algorithm follows in part the procedure described by Molavi et al.,Physiol Meas, 33, 259-270 (2012).

Inputs :

• data_dod - SNIRF data structure data, containing delta_OD data.
• mlAct - Cell array of vectors, one for each time base in data_dod, specifying active/inactive channels with 1 meaning active, 0 meaning inactive.
• iqr - parameter used to compute the statistics (iqr = 1.5 is 1.5 times the interquartile range and is usually used to detect outliers). Increasing it, it will delete fewer coefficients. If iqr<0 then this function is skipped.
• turnon - Optional argument to enable/disable this function in a processing stream chain.

Outputs :

• data_dod - SNIRF data structure containing modified delta_OD data size as dod (Channels that are not in the active ml remain unchanged).

Usage Options :

• Wavelet_Motion_Correction: dod = hmrR_MotionCorrectWavelet(dod, mlAct, iqr, turnon)

Default Parameters :

• iqr=1.50
• turnon=1

Prerequisites :

• Intensity_to_Delta_OD: dod = hmrR_Intensity2OD( intensity )