Wavelet-Based Statistical Analysis versus SPM of Brain Imaging Data
Radu Mutihac · International Journal of Intelligent Computing in Medical Sciences & Image Processing · 2008
Analysis of functional magnetic resonance imaging (fMRI) data of a block-based visual stimulation paradigm was comparatively performed by the discrete wavelet transform (DWT) in the wavelet domain and statistical parametric mapping (SPM) within the framework of the general linear model (GLM) [1]. The link is supported by the low-pass analysis filter of the DWT that can be similarly shaped to a Gaussian filter in SPM and by the subsampling scheme that provides means to define the number of coefficients in the low-pass subband of the wavelet decomposition [2]. Functional data processing in the wavelet domain was carried out by means of two biorthogonal transforms resulting in activation patterns similar to the activation maps obtained in SPM.