• Functional MRI
• Cerebral vasoreactivity
• Artery stenosis
• 3 Tesla

PURPOSE
In the clinical practice cerebral vasoreactivity (CVR) can be evaluated by Trans-Cranial Doppler (TCD) examination or by Nuclear Medicine. In this study we test the feasibility of Functional Magnetic Resonance Imaging (f-MRI) to measure the CVR in patient with stenosis of internal carotid artery (ICA) or middle cerebral artery (MCA). The results were compared to those obtained by TCD.
MATERIAL AND METHODS
We enrolled 21 patients affected by stenosis of the internal carotid artery (ICA) or the middle cerebral artery (MCA). All patients underwent a 3 Tesla MR examination. Cerebral vasoreactivity was evaluated by a breath-hold f-MRI protocol. F-MRI data analysis was performed using FEAT tool of FSL package. Activation Maps in Regions of Interest (ROIs) of the MCA were obtained for each patient, with calculation of the percentage of activated voxels (% Act Vox) and the maximum (Z-max) and mean (Z-mean) values of the magnitude of the response. BOLD indexes (maximum peak, slope half-maximum, full-width half-maximum) were also obtained by placing ROIs in the cortex of temporal lobe, the occipital lobe, the temporal-occipital junction and in the central fissure.
The MR protocol also included MR-angiography and MR perfusion sequences.
A breath-hold TCD examination for the evaluation of cerebral vasoreactivity was performed in 12 of 21 patients.
A ROC curve analysis was performed to compare the accuracy of breath-hold FMR with respect to breath-hold TCD in detecting a reduced CVR.
RESULTS
CVR evaluated by breath-hold f-MRI was significantly reduced in the cerebral tissue irrorated by the stenotic vessels. The group of patients with occlusion/subocclusion had a BOLD-CVR significantly reduced with respect to the group with moderate stenoses. Perfusional data showed a moderate positive correlation between the CBF and the BOLD-CVR.
The ROC curve analysis demonstrated that breath-hold FMRI has an accuracy of 80% in detecting compromised CVR as identified by TCD.
CONCLUSIONS
fMRI with breath-hold paradigm is an alternative tool in the evaluation of cerebral vasoreactivity. Its advantages are the non-invasiveness, the operator-indipendence and the possibility to perform in the same examination also morphologic and angiographic sequences for a complete evaluation of brain parenchima and circulation. The measurement of BOLD response could also be helpful in the assessment of cerebral vasoreactivity in the subjects in which the trans-cranial Doppler examination is doubt, not diagnostic or not practicable.