Functional brain imaging using EPI at 3 T

Abstract

Functional magnetic resonance imaging (fMRI) is becoming an important tool in the mapping of brain activation. However there are two main concerns that need to be answered before functional imaging can be considered truly useful as a neurophysiological tool. The first is that the detected activation may be derived from large veins and, thus, be spatially separate from the underlying brain activity. The second is the incomplete understanding of the brain transfer function and its relation to brain activity, blood flow, and metabolism. This work contains initial results that will help address these points. Models of the brain vasculature predict that signal changes on SE (spin-echo) images are expected to be much smaller in magnitude but very accurate in localizing true areas of activation than on GE (gradient-echo) images which are susceptable to large veins. By comparing activation from SE and GE EPI at 3 T, we have shown that the regions of activation are spatially very similar, suggesting that GE activation is closely linked to the underlying brain activity. We have identified an experimental impulse response of the brain following 8-s visual stimulation. This impulse response can be used to successfully predict the frequency response obtained experimentally and its shape suggests a resonance phenomenon. This suggests the brain transfer function can be modeled from linear response theory corresponding to the inherent feedback control mechanisms of the brain homeostasis. Continuation of this early work will help to identify the links between fMRI signal change and underlying brain physiology.