Poster

1
New methods for assessing the control of blood
flow in the brain
Hesam Kouchakpour, Supervisor Dr D.M Simpson
Prof, R. Allen ISVR, University of
Southampton
Introduction
Technical Information methodology

• Autoregulation is the automatic adjustment of
  blood flow to supply the required oxygen and
  glucose to each tissue in the body in proportion
  to the tissues requirement at any instant time.
  In other words for the brain, cerebral
  autoregulation is an active process of the brain
  by which cerebral blood flow is controlled at
  steady state despites the changes in the arterial
  blood pressure to ensure the required supply of
  the blood for the cerebral tissues.
• CA attracts considerable attention as it is
  thought to be an important mechanism in the
  development of some strokes, and also in the
  occurrence of the secondary damage, following
  stroke. The physiological control system is
  highly complex and is not fully understood in
  spite of extensive research that has been carried
  out in this area.
• In order to understand the challenges behind
  autoregulation ,it is necessary that the
  mechanism of the cerebral autoregulation is
  tested.
• Over the last two decades CA has been evaluated
  by measuring relative cerebral blood flow (CBF)
  response to a steady-state (static) changes in
  the arterial blood pressure (ABP)or the response
  to a sudden and rapid change in the blood
  pressure (dynamic response)
• In the most recent work dynamic autoregulation
  (dCA) has been used.
• The aim of this study is to propose innovative
  experimental and signal analysis techniques for
  the robust assessment of cerebral blood flow
  control

• As mentioned before the aim of this study is to
  propose innovative experimental and signal
  analysis techniques for the robust assessment of
  cerebral blood flow control. This is aimed to be
  done by
• explore new and innovative experimental protocols
  that exploit pseudo-random binary sequences in
  provoking variations in ABP and pCO2
• develop and evaluate advanced signal analysis
  procedures, that allow the dynamic interaction
  between CBFV, ABP and pCO2 to be quantified
• recommend sensitive and robust procedures for the
  non-invasive measurement of the blood flow
  control system in vulnerable patients
• increase the understanding of the dynamic
  interaction between CBFV, ABP and pCO2, based on
  modeling of experimental data

Time Frequency Domain Comparison
Autoregulation

• Four healthy adult volunteers blood pressure and
  cerebral blood flow were measured for the
  duration of around 50 minutes with the subjects
  at rest. During this time, volunteers breathed
  ambient air and air enriched with 5 CO2 through
  a face-mask. The latter provokes hypercapnia,
  known to temporarily impair autoregulation. The
  time series of spontaneous
• arterial blood pressure and cerebral blood flow
  velocity were collected with Finapres and
• transcranial Doppler ultrasound devices,
  respectively.
• Autoregulation can be estimated from the gain or
  phase-lead in the frequency response at around
  0.1 Hz. Coherence can also provide an estimate of
  autoregulation ( high coherence means absence of
  active autoregulation).

• The Basic system is having blood pressure as the
  input and the fluctuation of the cerebral blood
  flow as the output with the assumption that the
  cerebral blood flow is unaltered by a change in
  the cenrtarla arterial blood pressure and intact
  autoregulation would maintain the cerebral blood
  flow constant but it will not be able to
  maintain it if it is impaired.
• In classical autoregulation, there are three
  stages associated with the measurement of the
  cerebral blood flow with respect to changes to
  arterial blood pressure.. These stages can be
  seen in Figure 1.
• At very low and very high pressures where
  autoregulation is not active, cerebral blood flow
  will change with arterial blood pressure, and
  there is an intermediate level in the middle. In
  this plateau region the autoregulation is said to
  be active and changes in blood pressure will not
  alter changes in cerebral blood flow.

Figure 2 Step response from Frequency estimate
for both normocapnia and hypercapnia. (Bold line
is for High CO2 and dashed line is for normal CO2
levels).
Figure 3 Gain from Frequency domain estimate
Figure 1 Relationship between Arterial Blood
pressure and Cerebral blood Flow assuming
classical autoregulation
Dynamic Autoregulation (dCA)
Autoregulation comparison for normal (left side
of the lines) and high CO2 (right side of the
lines) levels for volunteers 1.Phase from
Frequency, 2.Phase from Time, 3.Gain from
Frequency, 4.Gain in Time
Figure 4 Gain from Frequency domain estimates

• Most recent work on cerebral autoregulation has
  focused on the transient response, known as
  dynamic cerebral autoregulation (dCAR). It has
  been shown that dynamic and static autoregulation
  have significant correlation for healthy human
  subjects (Aaslid, 1989). dCA can be quantified
  from the relationship between ABP and cerebral
  blood flow (CBF) through the use of appropriate
  signal processing methods.
• This can even be carried out even in the presence
  of only spontaneous variations of blood pressure
  which is clearly ideal as it avoids major
  interference with the patient. However, none of
  the methods of estimating dCA have been found to
  be sufficiently robust to be considered a gold
  standard nor have they been used routinely in
  clinical practice. Thus more research on
  reproducibility and method comparison is urgently
  needed in this field.

Conclusions and Outlook

• As can be seen from above figures, results from
  time- and frequency domain methods follow the
  same trend but with considerable scatter there
  is also clear bias in phase estimates with
  time-domain results being lower.
• Phase is again found to provide a stronger
  distinction between the experimental protocols,
  but large inter-individual variability makes it
  difficult to determine thresholds for
  normal/impaired autoregulation
• Time and frequency analysis are almost the same
  for time- and frequency domain approaches (SNR
  4.1673 and 4.3159,respectively).
• The small size of the sample analyzed, the short
  segments of recording , available and the absence
  of a gold standard measurement do no currently
  allow strong inferences regarding which method
  should be preferred.
• In future work, non-linear, time-varying and
  multivariate models (with CO2 variations as an
  additional input) will be investigated. New
  experimental protocols, in which higher
  variations in blood pressure are induced, will
  also be studied, with the aim of increasing the
  robustness of autoregulation estimates

1 Aaslid, R. K.-F. (1989). Cerebral
autoregulation dynamics in humans. 2 David M.
Simpson, 1. R. A Parametric Approach to Measuring
Cerebral Blood Flow Autoregulation
from Spontaneous Variations in Blood Pressure. 29.
References