 |
 |
Hi everyone, We are analyzing resting state fMRI data using the temporal concatenation version of ICA. The resulting components seem spatially sound, but within a couple of them, some of the subject mode values are negative while others are positive (within the same component). From what I’ve read on this forum I conclude that this is not an issue for spatial purposes. But I’ve been asked to correlate the subject modes to cognitive task scores acquired outside the scanner and I’m puzzled by these negative values. Christian’s response in the previous post worries me when he mentions that these modes are “not very interpretable” coming from resting state and I don’t fully understand the solution he offers to Philipp. Somewhere else on this forum it has been suggested to use the absolute values of the subject modes for correlation purposes, but that makes me wonder about the group contrasts. So I would like to pose the following questions: a) Am I on the wrong track interpreting subject modes from resting state data at all, and specifically in correlation with neuropsychological measures? b) If they are suitable, should given or absolute values of the subject modes be used? c) Are the group contrasts based on these subject modes? If so, are they based on the given or the absolute values? Many thanks in advance for anyone being so kind as to explain this to me. Ilse Schuitema