I intend to identify brain regions using spatial independent component analysis, as described in Bartels and Zeki (2004) and have a questions about an early step in this process. I would like to convert my preprocessed data into a large matrix with the number of rows corresponding to the number of scans (approx. 800-900) and the number of columns corresponding to the number of voxels (between 60,000 - 70,000). This matrix I will then submit to further processing. My questions is how do you take preprocessed data in SPM and convert it into a matrix? <div>

<br></div><div>I have done essentially this in Marsbar, for specific ROIs, so I imagine this would be a similar process, but for the whole brain instead. I would rather not use Marsbar if there are functions built into SPM that I can use. Any advice to get started would be greatly appreciated.</div>

<div><br></div><div>- Patrick B. Williams</div><div>Claremont Graduate University</div>