Empirical hemodynamic models of micro- and macro-vessels in the brain: an fMRI and RespirActTM approach

Functional Magnetic Resonance Imaging (fMRI) is the most-widely used technique
to study brain function non-invasively in humans for neuroscience and in the
clinic. fMRI measures brain function indirectly via local hemodynamic changes
that arise from neuronal activation. Despite the wide application of fMRI, a
key confound in the interpretation and quantification of the data is that the
signals measured consist of a mix of hemodynamic changes, some of which relate
to the vascular organization only (e.g. passive blood flow in macro-vessels),
and some of which relate to the metabolic needs of active neurons (e.g. blood
oxygenation changes in micro-vessels). The challenge of separating neuronally
driven from purely vascular signals has inspired significant model-development
work, primarily based on animal research. Translation of animal-based models to
humans is however problematic due to differences in vascular anatomy and
organization between species. Models tailored to the human brain will be
constructed by empirical modification of animal-based models. Separation of
neuronally driven and vascular signals can be achieved via reproducible
inspiration of CO2 and O2 during fMRI. CO2 and O2 inspiration affects vascular
signals independent of neuronal activity and is typically used to assess the
reactivity of the brain vasculature. Delivery of CO2 and O2 will be controlled
digitally by continuous analysis of expiratory gas and model controlled
inspiration of O2 and CO2 to achieve a reliable and repeatable gas delivery
using the RespirActTM. The measurements will be performed at 3 Tesla (T) and
7T. The vascular source of the fMRI signal is different at 7T as compared to 3T
due to the magnetic properties of blood and increased signal-to-noise. Detailed
7T measurements will link to the detailed analysis and models available from
animal research, and will in turn be used to translate models to the widely
available 3T scanners. The models will be tested in a separate group of
subjects for a range of neuronally driven signals without respiratory
challenge.

The main objective of the study is to characterize, quantify, and model
hemodynamic changes of different vessel groups in the human brain in response
to external stimulation.

Subjects will be included in a descriptive study for hemodynamic models of the
MR signal. The study involves two parts involving two subject cohorts. Part I:
fMRI with one simple task and respiratory challenge (n=28), part II: fMRI with
several simple tasks without a respiratory challenge (n=28). In part I,
parameters describing hemodynamics due to vascular (respiratory challenge) and
neuronal (simple task) processes will be extracted and modelled. In part II,
the parameters modelled will be tested for several simple tasks.

No benefits are expected for the volunteers. There are no known risks
associated with fMRI acquisition and the burden can be considered minimal. Two
or three visits will be required per subject.
Respiratory challenge: The controlled gas breathing requires a closed breathing
system and subjects have to breathe through a mask. Further, subjects may
experience discomfort due to the increased levels of CO2 in the blood. However,
end-tidal CO2 partial pressure levels of 40-57 mmHg (+5 - 10 mmHg above the
subjects* baseline value) are in physiological ranges and are experienced
repeatedly by most people over the day and during exercise. If subjects
experience discomfort because of high CO2 levels, they can open a valve in the
mask or we can switch immediately to 100% oxygen by pushing the red button on
the front of the gas blender. In the MRI, they can squeeze the emergency
button. Subjects can terminate the study at any time by communicating with the
operator. Approximately 6.5% (11 out of 168) of the subjects may experience
breathing discomfort during CO2 inhalation and terminate the study, and
approximately 8.9% (15 out of 168) of the subjects may experience breathing
discomfort and claustrophobia during CO2 inhalation and terminate the study, as
reported in previous studies conducted at the University Medical Center (UMC)
Utrecht. Risks associated with controlled gas breathing of high levels of CO2
with the RespirActTM are minimal because the minimum O2 level in all the gas
mixtures is 10%. The blood gas is controlled by a digitally controlled gas
blender steered by end-tidal blood gas pressure CO2 and O2. So, CO2 and O2
levels and breathing frequency are monitored online. An independent blood
oxygen saturation and respiratory rate monitoring will be performed during the
MRI measurements with fingertip pulse oximetry and a pressure sensor for the
respiratory rate. All staff that will operate the RespirActTM will be trained
by an experienced operator certified by the manufacturer.