Early Detection of Central Hypovolemia in Humans

In healthy subjects cerebral autoregulation maintains cerebral blood flow to
prevent the occurrence of hypo- or hyperperfusion of brain tissue. Cerebral
autoregulation causes vasoconstriction as cerebral perfusion pressure increases
and vasodilatation as cerebral perfusion pressure decreases. However, cerebral
autoregulation has its limits, and when the lower limit is reached, cerebral
blood flow declines further with development of symptomatic cerebral
hypoperfusion. Currently, volume status in anesthetized patients is monitored
by heart rate and mean arterial pressure. These parameters are not capable to
detect a subclinical hypovolemic state common in the perioperative period.
Thus, the application of these parameters to guide fluid therapy may result in
unrecognized hypovolemia or hypervolemia and a fundamental problem is that
detecting a volume deficit is not straightforward. More meaningful parameters
are desired but require monitoring of many signals, which are too complex to
interpret for the clinician. Complex software models based on physiological
data are needed to be able to evaluate and detect changes in the magnitude of
the central blood volume more adequately. A way of doing this is with a machine
learning model which takes into account multiple parameters.

To detect incipient central hypovolemia preceding symptomatic cerebral
hypoperfusion by machine learning models trained on realistic physiological
models.

Subjects are exposed to progressive central hypovolemia (by passive head-up
tilt, lower body negative pressure (LBNP) and active standing. Waveform
characteristics of continuous blood pressure, cerebral blood flow velocity and
non-invasive applanation arterial carotid and femoral pulse waves are extracted
and be used as input in a set of machine learning models to be trained on these
data to detect relevant biophysical signals related to cerebral hypoperfusion
in multiple timeframes.

There are no foreseen risks with participating in this study. The burden for
the subject is minimal because all measurements done in this study are
non-invasive. The physical load of the routine tests in this study is generally
well tolerated. During the entire study, the subject will continuously be
monitored to ensure subjects safety.

However, there is always the possbility of reaching syncope and fainting when
inducing pre-syncope. This will be carefully monitored and immediately
reveresed will this occur.