Content uploaded by Daniel J Licht
Author content
All content in this area was uploaded by Daniel J Licht
Content may be subject to copyright.
Post-Surgical Cerebral Autoregulation in Neonates with
Congenital Heart Defects Monitored With Diffuse
Correlation Spectroscopy
Erin M Buckley1, D A Goff2, T Durduran3,4, M N Kim1, R C Mesquita1, G Hedstrom5, D J Licht5, A G Yodh1
1Department of Physics and Astronomy, University of Pennsylvania, Philadelphia, PA 19104
2Division of Cardiology, 5Division of Neurology, Children's Hospital of Philadelphia and Hospital of the University of Pennsylvania,
Philadelphia, PA 19104
3ICFO- Institut de Ciències Fotòniques, Mediterranean Technology Park, 08860 Castelldefels (Barcelona), Spain
4Department of Radiology, Hospital of the University of Pennsylvania, Philadelphia, PA 19104
E-mail: ebuckle2@sas.upenn.edu
Abstract: Following cardiac surgery, cerebral blood flow changes in neonates with congenital heart defects are
measured using diffuse correlation spectroscopy. Using statistical correlations with mean arterial pressures, we
explore an “autoregulation index” to define periods of impaired autoregulation.
© 2010 Optical Society of America
OCIS codes: (170.3880) Medical and biological imaging, (170.5380) Physiology, (170.1610) Clinical applications.
1. Introduction
Approximately 6-8 in 1,000 infants born each year are diagnosed with congenital heart defects (CHD), a third of
whom require major surgical repair in the first month of life. Recent advances in cardiac surgery for severe CHD
have minimized infant mortality. Thus, clinicians are now focusing on the prevention of neurologic injury and the
improvement of neurocognitive outcome in these high-risk infants. Studies conducted with magnetic resonance
imaging (MRI) indicate that the incidence of neurological injury increases substantially after cardiac surgery[1, 2].
We hypothesize that disruptions of cerebral autoregulation in the post-operative period may contribute to brain
injury in these infants. Cerebral autoregulation, loosely defined as the ability to maintain adequate and constant
cerebral blood flow (CBF) despite changes in systemic blood pressures, is very difficult to measure non-invasively
due to limitations in CBF measurements. To assess hemodynamic autoregulatory changes, researchers have used
transcranial Doppler measurements of blood flow velocity in the main cerebral arteries, near infrared spectroscopy
(NIRS) measurements of cerebral blood volume, tissue oxygen saturation, and/or hemoglobin difference as a
surrogate for CBF. Using NIRS, Brady et al. have pioneered a real-time technique to assess autoregulatory failure
using measures of tissue oxygen saturation (StO2) [3]. They compute a 300 second moving window Pearson's
correlation coefficient between arterial blood pressure and StO2, coined the cerebral oximeter index (COx), which
they update every 60 seconds. Using a predefined threshold for COx, they have had success in both animal and
human models in identifying time periods of impaired autoregulation, i.e. times in which systemic blood pressure
and CBF exhibit a strong positive correlation [4]. The main weakness of this technique is the use of StO2 to infer the
behavior of cerebral blood flow instead of monitoring CBF directly.
Following the spirit of Brady et al.'s methods, we propose the use of diffuse correlation spectroscopy (DCS), a
novel, non-invasive optical technique capable of directly measuring relative changes in CBF in a continuous fashion,
as a means to assess a cerebral autoregulation index in neonates following cardiac surgery. DCS [5] has shown
promise as a monitor of relative changes in cerebral blood flow (rCBF) [6, 7]. Like NIRS, DCS employs near-
infrared light to probe the dynamics of deep tissues. However, DCS detects changes in CBF directly by monitoring
temporal fluctuations of scattered light. It does not rely on tracers or measurements of tissue oxygenation to
indirectly infer information about CBF. Finally, in contrast to Doppler ultrasound, DCS provides information about
microvascular as opposed to macrovascular hemodynamics.
To date six subjects diagnosed with complex congenital heart defects, either hypoplastic left heart syndrome
(HLHS) or transposition of the great arteries (TGA) have been recruited and monitored with DCS in the cardiac
intensive care unit for twelve hours immediately following return from cardiac surgery. Herein we describe our
optical and vital signs data acquisition and analysis techniques, and we employ DCS to quantify duration and
magnitude of impaired autoregulatory capacity in these neonates using the methods outlined in [3].
2. Procedure
With institutional review board approval, all newborn infants with CHD admitted to the cardiac intensive care unit
(CICU) at The Children’s Hospital of Philadelphia (CHOP) were evaluated for study inclusion. Inclusion criteria
included full term age (gestational age 40±4 weeks), an intention to undergo surgery with cardiopulmonary bypass
(CPB) with or without deep hypothermic circulatory arrest (DHCA) and medical stability for 24 hours prior to
surgery. Infants were excluded if they had a history of neonatal depression (5 minute APGAR<5 or cord pH<7.0) or
a pre-operative cardiac arrest requiring chest compressions. Subjects with HLHS treated with hypercarbia for
pulmonary over-circulation were also excluded.
On return from cardiac surgery, vital signs monitors were secured and the optical probe was placed on the
forehead. For data analysis purposes, vital signs, including heart rate, systolic, diastolic, and mean arterial blood
pressure, temperature, respiration rate, blood oxygen saturation, and right atrial pressure, were recorded via Moberg
Devices Component Neuromonitor Systems at a rate of 0.5 Hz. Manual time-stamps were placed to co-register the
optical and vital sign data. A hybrid NIRS/DCS optical device constructed in our lab (described previously [8])
acquired oxygenation and cerebral blood flow data at a rate of 0.13 Hz. The optical probe was repositioned on the
forehead routinely every two hours, or as needed in the case of poor signal quality or movement of the child. All
patients were monitored for 10 to 12 hours after returning from surgery.
The optical probe consisted of two source-detector pairs, both separated 2.5 cm apart, one designated for NIRS
measures of oxygenation, the other for DCS measures of rCBF. With this unique population in mind, we designed
the probe to be flexible and only 2mm thick. It was secured to the forehead with an eye mask. Anatomical MRI
scans reveal a combined skull, scalp, and cerebral spinal fluid thickness of 0.74±0.10 cm, thus the optical techniques
were probing a region between 0.8-1.2 cm into the cortex.
3. Data Analysis
For all optical data analysis, we assumed a semi-infinite geometry. For NIRS, we calculated changes in oxy-, deoxy-
and total hemoglobin concentrations using the modified Beer-Lambert law. Additionally, we quantified changes in
the absorption coefficient of the tissue, and we incorporated these changes into our DCS analysis. For DCS, we used
solutions of the electric field correlation diffusion equation to derive relative changes in cerebral blood flow. All
relative changes were calculated by subtracting/dividing by the mean during the first five minutes after probe
placement or readjustment.
Once relative changes in CBF and oxygenation were established, statistical correlation calculations were
examined and cerebral autoregulation was assessed. For simplicity, we present only the correlations between mean
arterial pressure (MAP) and rCBF. Following the steps laid out by Brady et al., the entire time course of data (rCBF
and MAP) was averaged over nonoverlapping 10-second intervals. A Pearson's correlation coefficient, dubbed the
cerebral blood flow index (CBFx), was computed from consecutive, paired, 10-second means of rCBF and MAP
from a 300 second window (30 data points). This 300 second window was then shifted 60 seconds and the CBFx
was recalculated from the new overlapping time frame. An arbitrary CBFx threshold of 0.5 was assigned to
designate impaired autoregulation.
4. Results
Data from six patients have been fully analyzed. Figure 1 shows samples of rCBF, MAP, and CBFx versus time for
a small portion of the total 12-hour monitoring. The figure demonstrates rCBF following the trend of MAP around
3, 15, and 25 minutes, at which time CBFx rises, suggesting possible impaired cerebral autoregulatory abilities.
Trends such as these are observed periodically throughout the monitoring period. To assess the extent of impaired
autoregulation in each subject, the percentage of time spent above the CBFx threshold was also calculated (Figure
2). Subjects spent a median(range) of 4.9(17.6) percentage of time with CBFx above the threshold.
5. Conclusions
This abstract described the first steps of a novel application of measures of relative changes in cerebral blood flow
obtained with diffuse correlation spectroscopy to assess cerebral autoregulation in infants with congenital heart
defects following cardiac surgery. More patients and correlations with concurrent measurements of oxy- and deoxy-
hemoglobin concentrations will permit further exploration of the utility of this new index.
6. References
1. Mahle, W.T., et al., An MRI Study of Neurological Injury Before and After Congenital Heart Surgery.
Circulation, 2002. 106(90121): p. I-109-114.
2. Galli, K.K., et al., Periventricular leukomalacia is common after neonatal cardiac surgery. Journal of Thoracic
and Cardiovascular Surgery, 2004. 127(3): p. 692 - 704.
3. Brady, K.M., et al., Continuous time-domain analysis of cerebrovascular autoregulation using near-infrared
spectroscopy. Stroke, 2007. 38(10): p. 2818-25.
4. Brady, K.M., et al., Continuous measurement of autoregulation by spontaneous fluctuations in cerebral
perfusion pressure: comparison of 3 methods. Stroke, 2008. 39(9): p. 2531-7.
5. Boas, D.A., L.E. Campbell, and A.G. Yodh, Scattering and Imaging with Diffusing Temporal Field
Correlations. Physical Review Letters, 1995. 75(9): p. 1855-1858.
6. Kim, M.N., et al., Noninvasive Measurement of Cerebral Blood Flow and Blood Oxygenation Using Near-
Infrared and Diffuse Correlation Spectroscopies in Critically Brain-Injured Adults. Neurocrit Care, 2009.
7. Zhou, C., et al., Diffuse optical monitoring of hemodynamics in piglet brain with head trauma injury. Journal of
Biomedical Optics, 2009. 14(3): p. 034015.
8. Buckley, E.M., et al., Cerebral Hemodynamics in Preterm Infants During Positional Intervention Measured with
Diffuse Correlation Spectroscopy and Transcranial Doppler Ultrasound. Optics Express, 2009. 17(15): p. 12571-
12581.
Figure 1 Sample time series of CBFx, MAP, and rCBF.
Figure 2 Percent of time in which CBFx was above our predefined threshold of 0.5.
