Impaired hemodynamic response in the ischemic brain assessed with BOLD fMRI.

The study aims to investigate the effect of cerebral ischemia or hypoperfusion in the evaluation of neural activity with blood-oxygen-level dependent (BOLD) functional magnetic resonance imaging (fMRI), and to examine whether the severity of the compromised hemodynamic status in patients with major cerebral artery diseases could, conversely, be assessed with the use of neural activity as endogenous vasodilator. 28 neurological impairment-free patients with anterior-circulation-territory ischemia performed a bimanual hand-grasping task. Magnitude and temporal shift of evoked BOLD response, baseline cerebral blood flow (CBF) and its increment, and the severity of hemodynamic impairment stratified by blood flow pattern were evaluated. For fMRI data, both conventional analysis with a canonical HRF and an HRF-model-free analysis were performed. The severity of hemodynamic impairment was significantly correlated (p<0.0001) with baseline CBF, CBF increment, and magnitude and delay of BOLD response. BOLD response delay was also significantly correlated (p<0.0001) with baseline CBF, CBF increment, and response magnitude. In 10 out of 45 ischemic motor cortices, conventional analysis completely failed to detect areas of activation that were demonstrated by HRF-model-free analysis. These data suggest that delay and reduced magnitude of BOLD response can be an indicator of the severity of compromised hemodynamic status, and that reduced regional baseline CBF and its increment underlie impaired BOLD response, which necessitates an alternative approach to conventional analysis with any single HRF.