Dear FSL-experts,
I am working on a quantitative MRI (T1-relaxometry) dataset consisting of two patient groups (with an unequal amount of subjects per group) that were scanned up to three times. Depending on the well-being of the patient, we could not conduct all three scans in all patients and therefore have to deal with some missing values. We read out T1 times in gray and white matter and found, employing a mixed model, significant group differences in T1 times adjusted for the latency of scan. We would now like to locate these group differences (and possible interactions with the latency of scan) using a voxelwise approach:
From what I read on the list, I understand that repeated measures are a bit tricky in randomise and I also stumbled upon PALM (though I have never used it before). Is there a possibility to set up a similar mixed effect model that can deal with our missing values in either one of those tools? And also, if this is possible, could you maybe give me a hint on how exactly this model and the respective contrasts can be set up?
Thank you already in advance, any help on this is greatly appreciated!
Greetings from Germany,
Bastian David
|