Dear SPMers,
I am trying to compare magnetization transfer ratio between two groups of patients one year after stroke. This is my first project using SPM so I am the very beginner. Could I ask you for advice whether this pipeline is correct?
Magnetization transfer ratio (MTR) is based on two sequences Mon and Moff, calculated according to the formula MTR=(Moff-Mon)/Moff x 100. MTR is basically a map of percentage units.
1. First, I reorient Moff and Mon using FLIRT in FSL because I was struggling with manual reorientation with SPM.
2. I coregister Moff and Mon
3. Calculate MTR in ImCalc using formula (i1-i2)./i1*100
4. I do segmentation and normalization of MTR and also Moff (Moff is partially T1 so I think this is better to get gray and white matter probability maps). To do this I use ALI (automated lesion identification) toolbox in SPM which is based on unified segmentation-normalization because it was designed for lesioned brains (such as after stroke in my case)
5. At this point, I have normalized MTR map, normalized gray and white matter probability maps taken from Moff.
6. To get rid of other tissue types I mask the normalized MTR map by gray and white matter in ImCalc i1.*((i2+i3)>0.2). I arbitrarily chose a threshold of 0.2.
7. Finally, I have MTR map of combined gray and white matter which I now smooth (kernel of 8mm)
Is this pipeline correct?
I would be grateful for any suggestions and comments.
Elzbieta Klimiec-Moskal
PhD Student
Jagiellonian University Medical College
Krakow, Poland
|