Transcript Advisory_board_meeti.. - University of Illinois at Chicago

Computer Assisted Design of Transport Processes in the Human Brain
Laboratory for Product and Process Design, Director A. A. LINNINGER
College of Engineering, University of Illinois,
Chicago, IL, 60607, U.S.A.
Grant Support: NSF, Medtronic, Susman and Asher Foundation.
Novel Imaging Techniques Transport & Kinetic Inversion
Motivation
• Millions of people are affected by
diseases of the Central Nervous System
(CNS)
visualize brain
functions
(E.g. Blood Flow
to pathological
organs)
structure of the
brain and NOT
its functions
• Systematic design of drug infusion
policies based on Transport and Kinetic
Inversion Problem (TKIP)
• Qualitative & Quantitative prediction of
treatment volume of site-specific drug
delivery from fluid mechanics
• Provide decision support to medical
community by specifying the parameters
for invasive drug delivery
fMRI – Used to
CT- Shows the
PET- detects
MRI- provides an
Schematic of BBB in the brain
• About seventy thousand people in U.S
are affected by hydrocephalus.
• Understanding pulsatile CSF dynamics
or intracranial dynamics is absolutely
necessary to predict and treat
hydrocephalus
Live Patient MRI
DTI- Used to
Cine MRI –
demonstrate the
structural properties
of anatomical
substructures
Flow velocities and
Cannot predict
intracranial
pressure and
tissue deformation
PET image of F-dopa-derived radioactivity, merged with magnetic
resonance image, computational grid and optimal result
0.09
0.08
•These advanced imaging techniques provide only
qualitative information.
• Non-invasive in-vivo MR
measurements cannot fully capture all
of the events of intracranial dynamics
• A quantitative first principles model is
presented that can accurately predict
patient-specific intracranial dynamics.
radioactive material
that is injected
or inhaled to produce
an image of the brain
anatomical view
of the brain
0.07
0.06
0.05
•Quantitative information such as drug diffusivity,
metabolic reaction constant, binding coefficient are
not directly available from these images.
0.04
0.03
0.02
0.01
Hydrocephalic Brain
Methodology
•Knowledge about these parameters is important in
systematic design of drug delivery policies.
0
Clinical concentration field of
L-dopa
Patient Specific Quantification of Intracranial Dynamics
CSF Flow Field during one cardiac cycle – Normal brain
MR Imaging
Brain Geometry, CSF flow field
Reconstruction tools
ImageJ, Insight SNAP, Mimics
Optimal result,
Computational grid
M eD
Quantification of CSF flow field
Direct Experimental
Measurements
Intracranial Pressure Pulsatility during one cardiac cycle – Normal brain
2D and 3D geometry
of the Ventricles and
Subarachnoid space
Quantification of Intracranial Pressure
Geometry
Grid Generation
Gambit
CSF Flow Field during one cardiac cycle – Hydrocephalic brain
Computational Mesh
Computational Fluid Dynamics
Continuity and Navier-Stokes Equations
for CSF Motion
Quantitative analysis
Velocity field and CSF dynamics
Prediction of Intracranial pressure (ICP)
Intracranial Pressure Pulsatility during one cardiac cycle – Hydrocephalic brain
Analysis of flow
And pressure
patterns
t= 0 %
t= 60 %
t= 30 %
Tissue Properties
t= 90 %
• CSF Pulsatility increases 2.3 times than normal in hydrocephalic case
CSF flow and ICP measurements from fluid mechanics
• ICP increases by a factor of four in hydrocephalic case
Quantitative Prediction of Drug Distribution
Boundary Conditions
Prediction of treatment volume in a 2D coronal cut of a human brain using NGF as drug
1st week
3rd week
2nd week
Regions of interest in targeted drug delivery
4th week
Estimation of Penetration Depth
Tissue properties
Site-specific drug delivery
Neurodegenerative
disease
Drug
GDNF
Parkinson’s Disease
Present Case Study
Drug: NGF
Target: Caudate Nucleus
Injection Location: 1. Thalamus
Infusion Site
Animal /
Human
IP: Putamen
Rhesus Monkey
Gash et al, 2003
IP: Substantia Nigra
Rhesus Monkey
Gash et al, 2003
IP: Putamen
Human Trials
Gill S.S et al , 2003
Ventricles & and Central
part of the Putamen
Rhesus Monkeys
Grondin R et al, 2002
ICV
Monkeys/Human
Trials
GDNF/ rmetHUGDNF
Parameter
Reference
Nutt J.G et al, 2003
Conclusions
φ–porosity
•Higher Treatment Volumes were realized
for high flow Infusion at the thalamus
Value
GM
WM
0.2
0.28
k – permeability GM
WM
10+16 m-2
X: 10+13 m-2
Y: 10+13 m-2
β – inertia resist. GM
WM
8.32*109 m-1
X: 5*108 m-1
Y: 2*108 m-1
Tortuosity
3
1.0
•The total treatment volume at the end of 4
weeks was found to be 0.107 cc
GM
WM
Acknowledgements
•Accurately reconstruction of the human brain geometry to quantify transport processes.
 Dr. Richard Penn, University of Chicago
•A novel method for extracting transport and reaction constants from experimental data was presented based on TKIP
 BRIC, University of Chicago
•Prediction of treatment volumes based on site-specific drug delivery for NGF was presented.
 Fluent Inc, Lebanon, NH
•Accurate quantification of CSF flow and Intracranial pressure fields.
 Materialise Inc, Ann Arbor, MI
• Validation of CFD simulations with Cine Phase MRI measurements at select regions of the ventricular system.
 ImageJ, NIH, MD.