This study measures the "tortuosity" or twistedness of brain arteries to determine if increased tortuosity is predictive of intracranial aneurysm development. Researchers will compare the tortuosity of 500 high-risk patients to normal-risk patients and establish a tortuosity risk indicator. Tortuosity will be measured using a "distance factor metric" that divides the actual path length of the artery by the straight line distance between endpoints.
2. Summary
• Intracranial aneurysms are balloon
like bulges in the artery walls in the
brain
• Ruptures cause strokes usually
resulting in severe disability or death
• This study measures the twistedness
or tortuosity of brain arteries to test
if increased tortuosity predicts
intracranial aneurysm development
2
3. Tortuosity & Risk
• Compare tortuosity of 500 patients
from known high risk pedigrees with
normal risk patients [Farnham, 2004]
• Establish tortuosity risk indicator
– High aneurysm risk Low aneurysm risk
–
– Expect high tortuosity Expect low tortuosity
3
5. Tortuosity Measure
• Distance Factor Metric = L/d
– Divide the distance L along
the centerline by the straight
end to end distance d.
– Use a bifurcation as a starting L/d
point and calculate L/d ratio
for every voxel along the
artery.
5
6. Maximum Intensity Projection (MIP)
• Rays through
image stack
• Keep maximum
intensity pixel
• Goal: extract
blood vessel
from rest of
image in 3-D
6
7. MIP Z-Buffer
• Intensity is
position in image
slice stack of
maximum pixel
intensity; dark is
closer, brighter is
farther
• Contiguous blood
vessels are
smooth
7
8. 3-D Region Growing
• Check if pixels
neighboring 26
voxels are above
seed histogram
threshold and add
non-maximal 3-D
pixels
• MIP of 3-D region
grown image
stack
8
9. Finding Centerlines
Highest
Neighborhood
modified
Distance From Edge (DFE) Modified DFE (MDFE)
Goal
Inverse Lowest
higher cost
MDFE paths to
is lower every voxel
Path Cost
cost Voxel weights/cost
Start from farthest voxel
from goal and work back
to goal or an earlier path
Centerline 9