Analisis Data Pengimejan Resonans Magnet Kefungsian: Pra Pemprosesan Ruang Menggunakan Kaedah Pemetaan Statistik Berparameter

Abstract

Baseline functional magnetic resonance imaging (fMRI) study has been carried out on 2 healthy male subjects (left and right handed), aged 22 and 25 years old, respectively. The fMRI scans were Performed using a 1.5 T magnetic resonance imaging (MRI) system at the Department of Radiology, Universiti Kebangsaan Malaysia Hospital. The study used the movement of the right- and left-hand fingers to stimulate neuronal activity in the cerebral cortices. A five-cycle active-rest paradigm was used with each cycle consisted of 1 active block and 1 rest block which individually consisted of 10 series of measurements. The fMRI images were analysed using MatLab and statistical parametric mapping 2 (SPM2) software packages. A rigid body registration using 6-parameter affine transformation was performed on all T2*-weighted functional images. The results showed that the subject's movement was minimum in either translational (< 1 mm) or rotational (< 1o) direction. All images were normalized via a nonlinear warping using a 12-parameter affine transformation and were found to match a template that already conform to a standard anatomical space. However, the shape, resolution and contrast of the functional images were slightly altered as compared to the originals. Image smoothing using an isotropic 6 mm Gaussian kernel rendered the image data parametric with a considerable loss in resolution and contrast. Structural segmentation performed on T1-weighted images classified brain tissues into grey matter, white matter and cerebrospinal fluid. The spatial preprocessing of the functional and structural images rendered the data parametric with Gaussian type of distribution, ready to be analysed using the general linear model and Gaussian random field theory.