 |
 |
Hi, just to add to the previous responses: 1) Consider getting a whole-head PD-image. Lack of flow void in the internal carotid artery indicates ACI occlusion or near-occlusion. 2) Consider ASL and try to calibrate to an M0-acquisition. Although highly dependent on the specific parameter settings, it may be valuable for perfusion quantification. Cheers- Andreas ________________________________ Von: FSL - FMRIB's Software Library im Auftrag von Julius Fridriksson Gesendet: Mo 15.08.2005 15:06 An: [log in to unmask] Betreff: Re: [FSL] fMRI stroke preprocessing/analysis Hi Carlos -- Heidi raises some very important issues to keep in mind with MRI and stroke patients. I want to add that we have had much better results using event related rather than block design in this population; especially in the case of abnormal cerebral perfusion. Good luck -- Julius >>> [log in to unmask] 08/15/05 04:48AM >>> Dear Carlos, Although there are lots of additional things that you might want to consider when scanning stroke patients, there is not a single standard approach to this and many people would simply use the same processing any analysis that you would use for studies of healthy controls. Having said that, it is worth thinking through the issues you mention, to decide whether you want to do anything special. For coregistration - if your patients have large cortical lesions then default coregistration approaches may not be appropriate. You may want to consider using cost-function masking, within FLIRT, or else using a non-linear registration. Alternatively, you could derive your activation statistics in native space using ROIs, and avoid the issue of coregistation. For structural imaging - depending on the timing and type of lesions in your patient you may want to consider acquiring different structural images for lesion characterisation- this really depends what you want to get out of the data. For BOLD modelling - although it is possible that the hemodynamic response is altered in patients, most studies ignore this possibility and simply use the same generic bold modelling as for a standard fMRI study. If you want to take into account the possibility that the shape of the hrf is altered in patients then you could use basis functions for more flexible response modelling. If you want to characterise the hrf in some detail to compare across groups then this would be easier done with a single event, randomised design than with a block design. Heidi Carlos R. Cortes wrote: >Hello Dear FSL users, > >I have been working with fMRI in Schizophrenia and ADHD for sometime (Just >dealing with EPI images), but now I am starting to work with people post- >stroke (> 6 months). This is my first week and I have a patient next week >(not enough time to read and implement changes), it would be great if >somebody give me some hints about: > >1) Considerations in preprocessing (e.g. DTI, FLAIR, T1 or T2 >coregistration). > >2) Considerations in analysis (BOLD modeling for patients) > >Any comment will be useful, > >Carlos > > -- Dr Heidi Johansen-Berg Wellcome Trust Training Fellow Oxford Centre for Functional Magnetic Resonance Imaging of the Brain John Radcliffe Hospital Headington Oxford OX3 9DU http://www.fmrib.ox.ac.uk/~heidi Tel: 01865 222782 Fax: 01865 222717 email: [log in to unmask]