Hello Kate,

It would really depend on you in a way...I don't think you would want to normalize your data to a standard space (such as MNI) because that might interfere with the localisation of the infarcts. In such a case, you might want to just run a few of the other preprocessing steps such as realignment and estimation of scrubbing parameters? If you, on the other hand, don't think that any preprocessing is neccessary for your data, you could skip straight to denoising. If you were to run the default pipeline, it would end up transforming your scans to the MNI space and the normalization module might end up even eliminating the infarct. If you really did want to work in the standard space, then you would have to block out the regions where the infarct is and apply that masking during the normalization step.


Hope this helps



Best
Pravesh