Ultrasound Guided Near Infrared Spectroscopy for Brain Functional Study: Feasibility Analysis and Preliminary Work

1
Ultrasound Guided Near Infrared Spectroscopy for
Brain Functional Study Feasibility Analysis and
Preliminary Work

• Mr. Bo Qiang, Dr. Ronald Xu, Dr. Jun Liu
• Biomedical Engineering Center
• 286 Bevis Hall
• The Ohio State University
• Tel [phone]
• Email xu.202_at_osu.edu

2
Physiological Events Accompanied with Brain
Activities

• Intravascular
• Cerebral blood volume
• Cerebral blood flow rate
• Cellular
• Changes in neurons membrane potential,
  electrical and magnetic field caused by water and
  ion fluxes (Na, K, Cl-, Ca2, )
• Glucose and oxygen consumption, flavoproteins,
  cytochrome-c-oxidase (CO)

Simulated Oxygen Deficiency during Stroke
Brain Functional Imaging by Near Infrared Light
Opportunities, Status and Challenges

• Opportunities
• The well known functional imaging modalities like
  fMRI has shown the change of oxy/deoxy ratio
  during brain activities.
• Near infrared light is sensitive to physiological
  changes in cortex layer
• Status
• Time Domain, Frequency Domain and Continuous Wave
  devices have been developed for brain functional
  study.
• Optical imaging modality has been combined with
  other imaging modalities (MRI, PET,).
• Challenges
• Low spatial resolution.
• CW devices, although less expensive, are not
  proper for quantitative measurement
• Bulky probe head, hair interference and motion
  artifact

Franceschini et. al 2000
3
Our Proposal Miniature NIR Sensor with
Ultrasound Guidance and Optical Baselines

• Develop a miniature NIR sensor (CW type, silicon
  photodiode detector) that sticks on patients
  head with built-in opto-electronics for
  continuous monitoring
• Quantitative measurement of the optical property
  changes in cortex layer during brain activities
• Brain optical property baseline is obtained by a
  TD/FD device in advance
• Brain structure baseline obtained from MRI/CT in
  advance, minor structural variation monitored by
  ultrasound in real time
• Target neonatal and infant applications

Hypothesis During brain activities, the optical
and mechanical properties of the head remain
relatively constant except the cortex layer
absorption coefficient.
4
Evaluate NIR Feasibility by Cortex Optical
Pathlength and SNR Analysis
cortex optical pathlength (cortex sensitivity)
Adjust position of detector
source
Monte Carlo Simulation
Human head 5-layer model
Xu et.al 2005
Conclusion To reach the same optical sensitivity
on adult model, detector needs to be more
sensitive and have larger dynamic range
Human head anatomy
5
Feasibility of Ultrasound Sensors to
Differentiate Layers on a Head Simulating Phantom
Conclusion Ultrasound can detect the boundaries
of different layers in a head simulating phantom.
Ultrasound echo can help reconstruct the actual
thickness of each layer
6
A 2 Layer Simplified Model for Inverse
Calculation of Cortex Absorption
Source
Detector
L
Parameters used in simulation
n ms (cm-1) ma (cm-1) Thickness T (cm)
Layer 1 1.4 10 0.1 0.2
Layer 2 (cortex) 1.4 10 0.1 0.3 Inf.
ms_1 , ma_1
Layer 1
Layer 2 (cortex)
ms_2 , ma_2
Forward calculation use Kienles approximation
to avoid heavy computation load of Monte Carlo
method
7
Inverse Algorithms Calculated vs. Actual Cortex
Absorptions
Neuron Network Approach
Optimization Approach
Both algorithms successfully reconstructed the
cortex layer absorption with accuracy under 10
white noise.
Plot of Relative cortex sensitivity
Design Source-Detector Separation for Maximal
Cortex Sensitivity and Minimal Spatial
Inconsistency
Optimal source detector separation for this model
is 2.15 cm for infant model
8
Conclusions

• Simulation and benchtop tests demonstrated
  feasibility of using a low cost NIR/Ultrasound
  system for infant brain functional study.
• Inverse algorithm has been developed for real
  time calculation of cortex layer absorption based
  on a 2 layer model
• Source-detector separation was optimized for
  maximal cortex sensitivity

Future Work

• Development of the portable NIR/Ultrasound sensor
  for brain functional study
• Concept validation on animal models and infants
  for NIR/Ultrasound measurement of cortex
  activities
• Development of inverse algorithm for real time,
  absolute measurement of cortex absorption on a
  multi-layered tissue model

Data analysis and interpretation
Real time brain activity monitoring system
Computer clusters
Wireless communication channel
Post-processing and monitoring