Registration motivation

advertisement
Registration motivation
Mads Nielsen
Registration in Medical Imaging

Find correspondences – intrapatient:
1
inhale phase to exhale phase
1
Castillo, R.,
Castillo, E., Guerra, R., Johnson, V.E., McPhail, T., Garg, A.K., Guerrero,
T. 2009 “A Mads
framework
for
Statistical
Network
270312
Nielsen
evaluation of deformable image registration spatial accuracy using large landmark point sets” Phys Med Biol 54 1849-1870.
2
Monitoring subtle changes
Local is better!
Baseline
Follow-up
Problem: which pixel goes where?
Statistical Network
270312 Mads Nielsen
Image registration
Transform
Transform Coefficients
Optimization
Incorporate model of
density change
Cost
Function
Statistical Network
270312 Mads Nielsen
Disease monitoring using
image registration
Possible to indicate which locations change
Consistent in time
Local changes predict decline better
Statistical Network
270312 Mads Nielsen
Registration in Medical Imaging

Find correspondences – intrapatient:
disease progression
2
2Marcus,
DS, Fotenos, AF, Csernansky, JG, Morris, JC, Buckner, RL, 2009. Open Access Series of Imaging Studies
(OASIS): Longitudinal
MRI Data in Nondemented and Demented Older Adults. Journal 270312
of Cognitive Neuroscience,
in press.
Statistical
Network
Mads Nielsen
6
Readings: Atrophy
Assessment
Step 1: Whole Brain
Segmentation on baseline
Step 2: Un-biased Multi-scale
Nonlinear Deformation from
Baseline to Follow-up
Step 3: Estimate the volume
change per voxel from the
deformation field
Step 4: Estimate the structure
change by summing up values in
Step 3
Statistical Network
270312 Mads Nielsen
7
Readings: Atrophy Accuracy
Test
101 Subjects From ADNI. Changes from BLM12
Group
CrossSectional
Synarc (%)
Registration
Based(%)
-0.8 ± 1.4
-0.6 ± 2.4
Hippocampus
baseline
volume
Synarc (mm3)
2333 ± 330 (L)
2268 ± 299 (R)
-1.9 ± 2.0
-1.4 ± 2.9
-0.8 ± 2.0
-0.9 ± 2.9
-0.5 ± 7.0
-1.2 ± 1.5
-1.4 ± 3.0
2081 ± 312 (L)
2075 ± 363 (R)
-2.5 ± 3.1
-3.1 ± 3.3
-1.1 ± 1.2
-4.3 ± 3.1
-4.0 ± 7.1
-4.2 ± 2.4
-4.4 ± 4.1
1787 ± 482 (L)
1773 ± 484 (R)
-4.0 ± 4.4
-5.2 ± 4.3
-4.2 ± 1.6
Hippocampus
baseline volume
FreeSurfer
(mm3)
3410 ± 329 (L)
3430 ± 416 (R)
CrossSectional
FreeSurfer
(%)
-0.9 ± 2.6
0.9 ± 6.6
Longitudinal
FreeSurfer
(%)
MCI
(N =
37)
2992 ± 281(L)
3070 ± 420(R)
AD
(N =
57)
2542 ± 349(L)
2728 ± 621(R)
Normal
(N =
27)
Statistical Network
-0.6 ± 1.9
-2.0 ± 2.0
-4.4 ± 2.4
270312 Mads Nielsen
8
Registration at Large and Small
Scale

Inter-patient variation at large scale
disease progression
Atrophy at
smaller scale
Data from: Marcus, DS,
Fotenos, AF, Csernansky, JG,
Morris, JC, Buckner, RL, 2009.
Open Access Series of Imaging
Studies (OASIS):
“Longitudinal MRI Data in
Nondemented and
Demented Older Adults.“
Journal of Cognitive
Neuroscience, in press.
2
Statistical Network
270312 Mads Nielsen
9
LDDMM/LDDKBM registration



Large Deformation Diffeomorphic Metric
Mapping
Domain Ω⊆ℝ d , deformations ϕ∈ GV , ϕ:Ω →Ω
Find ϕ minimizing
E(ϕ)= E 1 (ϕ)+ U (ϕ)


E1 (ϕ) regularization/smoothness term
U (ϕ)= U (ϕ. I m , I f )
matching term
Images
If
Im
Statistical Network
270312 Mads Nielsen
10
Manifold/Lie Group Formulation
in LDDMM, regularization E 1 is the length of minimal
paths
If Sobolev-norm <Lv,v>
on v, then diffeomorphism
L is the momentum
Operator: Lv = a
Vt = ∫K(:,x)at(x)dx
Statistical Network
270312 Mads Nielsen
11
Download