Prospettive e limiti dei metodi RM nello studio della funzionalita cerebrale

In the last decade, nuclear magnetic resonance has become one of the main imaging techniques exploring brain function, offering a non invasive tool with good spatial-temporal resolution, disclosing changes in dynamic and metabolic parameters linked to neuronal activation. These studies are based on the hypothesis that brain function is spatially localised and neuronal activation is strictly correlated to local changes in certain metabolic and haemodynamic parameters in the brain. Functional magnetic resonance imaging studies provide information on the brain areas activated by specific stimuli, acquiring fast images of the brain during sensory or cognitive stimulation. In particular, functional magnetic imaging based on BOLD (blood oxygen level dependent) contrast exploits the paramagnetic properties of deoxyhemoglobin which thus acts as an intravenous endogenous contrast agent. Changes in MR signal can be traced to local changes in deoxyhemoglobin content and hence the metabolism responsible for neuronal activation. In this context, however, a good understanding of the haemodynamic and metabolic mechanisms involved in activation is important since the degree of local blood oxygenation is concomitantly dependent on CBF (local blood flow), CBV (local blood volume) and CMRO₂ (speed of local oxygen consumption). Despite this, a general physiological model of neuronal activation is not yet available. The limitations of the BOLD technique include signal interpretation and spatial- temporal resolution while the specificity of activation maps is conditioned by vascularization and the metabolic-haemodynamic response correlated to activation. Further critical aspects of the method include the statistical analysis of fMRI data, especially the quantitative assessment of the statistical significance of regions deemed activated.