Biomedical Engineering Reference
In-Depth Information
Like the method based upon injected contrast agents, intrinsic contrast is
sensitive to T 2 *-weighted signal changes and critically depends on the observa-
tions of Pauling and Coryell (197,198) that the magnetic properties of hemoglo-
bin change from the oxy-state, which is diamagnetic, to the deoxy-state, which
is paramagnetic. Because of this issue, the intrinsic contrast technique has been
called blood oxygen-level-dependent contrast imaging or BOLD (190). Blood
oxygen-level-dependent contrast results from a set of effects initiated by
changes in local cellular activity. These effects include alterations in cerebral
blood flow (CBF) and cerebral blood volume (CBV) that in general produce
increased oxygen delivery beyond oxygen utilization, so that there is a relative
decrease in local deoxyhemoglobin concentration. A relative decrease in deoxy-
hemoglobin concentration results in an increase in the relaxation time T 2 *, or
apparent T 2 , leading to an increase in MR signal in brain regions with increased
neural activity.
The mechanistic details behind this general model of intrinsic contrast fMRI
continue to be a topic of active research, specifically around (a) the neural corre-
lates of BOLD and other perfusion-weighted signals, (b) the coupling of neural
activity with vascular responses, and (c) factors influencing the concentration of
deoxyhemoglobin. What follows is a synopsis of research on the first and last of
these topics, given their relevance to interpretation of fMRI studies of normative
motivation, and altered motivational function in the form of neuropsychiatric
illness.
In a number of circumstances, BOLD signal changes have been observed to
be proportional to changes in neuronal spike rates (213). But other work involv-
ing stimulation of parallel fibers with neutralizing effects on measured spike
rates has shown circumstances where spike rate and CBF diverge (166). Data
have been further presented that local field potentials (LFPs) correlate better
with CBF/BOLD effects than spike rate (156), suggesting that changes in BOLD
signal reflect incoming synaptic activity and local synaptic processing. This re-
lationship between LFPs and CBF/BOLD will vary with local neural architec-
ture in that this relationship has been observed to be linear during climbing fiber
stimulation, and nonlinear with parallel fiber stimulation (166).
When neural activity is altered, corresponding effects are observed in CBF,
CBV, and oxygen consumption. The weight accorded to these effects and their
impact on deoxyhemoglobin concentration has been the topic of intense investi-
gation. Early in the development of fMRI, capillary perfusion studies in the rat
brain demonstrated that essentially all cerebral capillaries are perfused in the
basal state, and that increases in perfusion are accomplished primarily by in-
creases in blood velocity within capillaries, as opposed to the capillary recruit-
ment that is such a large factor in muscle (255). This observation was important
in that it defines clear limits for the coupling ratios of blood flow to volume and
to oxygen utilization. Activation-induced changes in oxygen utilization
(CMRO2) are thought to be coupled to changes in blood flow by a diffusion
Search WWH ::




Custom Search