This is where I have started collecting the various Matlab-based tutorials and demo code/data that help build intuition about fMRI data acquisition and natural scene analysis.
Topic Code Sample data 7T sample data
Acquired 9/17/09 with the new volume Tx, 16 ch Rx coil. Paradigm in both cases was rotating wedges (retinotopic mapping). 10 1/2 24-s cycles (total scan duration 252 s). Designed for Fourier analysis at 10 cycles/scan after discarding the first 12s of data.Two acquisitions: 1.5mm (isotropic) voxels w/ 1.5s TR (32 slices, oblique coronal orientation, R/L phase-encode, iPAT=4), and 2mm (iso) resolution and 2s TR (oblique sagittal orientation, 45 slices, iPAT=2). cw_1p5mm.nii.gz (86MB)
cw_2mm.nii.gz (60MB)
DICOM: cw_1p5mm.tar.gz (88MB)Linear Discriminant Analysis
A work in process ...caoLDA.m Principle Components Analysis
Commonly touted as either a data cleaning or summary method, it is a one-liner in Matlab once you figure out how your matrices are organized.caoPCA.m slice.mat Fourier Transform
Indispensable in fMRI data acquisition and analysis, as well as vision research.caoFFT.m zebras.mat Phase Scrambling of Natural Images
The workhorse code (generating the von Mises distribution) is compliments of Paul Schrater. The tutorial code uses prsVonMisesRand.m and caoReflectPhaseSpectrum.m; caoPhaseScramble.m is stand-alone.caoPhaseScramble.m
sampleOutput.jpg
sampleOutput2.jpg
caoPhaseScramblingDemo.m
prsVonMisesRand.m
caoReflectPhaseSpectrum.m
Bandpass Noise
Code to generate normally distributed random noise in particular spatial frequency and orientation bands.caoLandyTexture.m
caoTextureDemo.m
