Medical Image Computing and Computer-Assisted Intervention - MICCAI 2015

Name: Medical Image Computing and Computer-Assisted Intervention - MICCAI 2015 | 18th International Conference, Munich, Germany, October 5-9, 2015, Proceedings, Part III
Brand: Springer
Price: 53.49 EUR
Availability: OnlineOnly

18th International Conference, Munich, Germany, October 5-9, 2015, Proceedings, Part III

Nassir Navab Joachim Hornegger William M. Wells Alejandro Frangi(Editor)

Springer (Publisher)

Published on 28. September 2015

XXVII, 780 pages

E-Book

PDF with digital watermarking

System requirements

978-3-319-24574-4 (ISBN)

€53.49incl. 7% vat

System requirements

for PDF with digital watermarking

E-Book Single Licence

Available for download

Description

More details

Other editions

Content

Intro
Preface
Organization
Contents - Part III
Quantitative Image Analysis I: Segmentation and Measurement
Deep Convolutional Encoder Networks for Multiple Sclerosis Lesion Segmentation
1 Introduction
2 Methods
3 Experiments and Results
4 Conclusions
References
Unsupervised Myocardial Segmentation for Cardiac MRI
1 Introduction
2 Related Work
3 Method
4 Results
5 Discussions and Conclusion
References
Multimodal Cortical Parcellation Based on Anatomical and Functional Brain Connectivity
1 Introduction
2 Materials
2.1 RS-fMRI Data
2.2 dMRI Data
3 Methods
3.1 Estimating Brain Connectivity
3.2 Adaptively Weighted Multimodal Connectivity Model
3.3 Affinity Matrix Estimation
4 Results and Discussion
4.1 Quantitative Results
4.2 Qualitative Results
5 Conclusions
References
Slic-Seg: Slice-by-Slice Segmentation Propagation of the Placenta in Fetal MRI Using One-Plane Scribbles and Online Learning
1 Introduction
2 Methods
3 Experiments and Results
4 Conclusion
References
GPSSI: Gaussian Process for Sampling Segmentations of Images
1 Introduction
2 Existing Generative Models of Segmentations
3 GPSSI
3.1 Definition
3.2 Efficient Sampling
4 Results
4.1 Parameter Setting
4.2 Segmentation Sampling
5 Tumor Delineation Uncertainty in Radiotherapy
6 Conclusion
References
Multi-Level Parcellation of the Cerebral Cortex Using Resting-State fMRI
1 Introduction
2 Methodology
2.1 Data Acquisition and Preprocessing
2.2 Initial Parcellation via Supervertex Clustering
2.3 Single-Level Parcellation via Hierarchical Clustering
2.4 Groupwise Parcellation via Spectral Clustering
3 Results
4 Conclusions
References
Interactive Multi-organ Segmentation Based on Multiple Template Deformation
1 Medical Motivation and Overview
2 Methodology
2.1 Multiple Implicit Template Deformation with User Constraints
2.2 Numerical Optimization
2.3 Enhancing the Framework for Local Contours Editing
2.4 Flexibility of the Framework
3 A Study for the Evaluation of the User Interactions
4 Conclusion
References
Segmentation of Infant Hippocampus Using Common Feature Representations Learned for Multimodal Longitudinal Data
1 Introduction
2 Method
2.1 Hierarchical Learn ning of Common Feature Representations
2.2 Patch-Based Label Fusion for Hippocampus Segmentation
3 Experimental Results
4 Conclusion
References
Measuring Cortical Neurite-Dispersion and Perfusion in Preterm-Born Adolescents Using Multi-modal MRI
1 Introduction
2 Methods
3 Results
3.1 Comparison of DWI with and without Additional T2 Imaging
3.2 Comparison of Quantitative Neuroimaging Parameters
3.3 Correlation of Diffusion MRI and Cerebral Blood Flow
4 Discussion
References
Interactive Whole-Heart Segmentation in Congenital Heart Disease
1 Introduction
2 Patch-Based Interactive Segmentation
3 Empirical Study: Active Learning for Reference Selection
4 Results
5 Conclusions
References
Automatic 3D US Brain Ventricle Segmentation in Pre-Term Neonates Using Multi-phase Geodesic Level-Sets with Shape Prior
1 Introduction
2 Method
3 Experiments and Results
4 Discussion and Conclusion
References
Multiple Surface Segmentation Using Truncated Convex Priors
1 Introduction
2 Method
3 Experimental Methods
4 Results
5 Discussion and Conclusion
References
Statistical Power in Image Segmentation: Relating Sample Size to Reference Standard Quality
1 Introduction
2 Derivation of the Sample Size Formula
2.1 Sample Size for Segmentation Accuracy
2.2 Sample Size in Terms of the High Quality Reference Standard
3 Simulations
4 Results
5 Case Study
6 Discussion
References
Joint Learning of Image Regressor and Classifier for Deformable Segmentation of CT Pelvic Organs
1 Introduction
2 Method
2.1 Joint Learning of Image Regressor and Classifier
2.2 Deformable Segmentation with Regressor and Classifier
3 Experimental Results
4 Conclusion
References
Corpus Callosum Segmentation in MS Studies Using Normal Atlases and Optimal Hybridization of Extrinsic and Intrinsic Image Cues
1 Introduction
2 Methods
3 Evaluation Results
4 Conclusions
References
Brain Tissue Segmentation Based on Diffusion MRI Using l 0 Sparse-Group Representation Classification
1 Introduction
2 Approach
3 Experiments
3.1 Data
3.2 Diffusion Parameters
3.3 Comparison Methods
3.4 Results
4 Conclusion
References
A Latent Source Model for Patch-Based Image Segmentation
1 Introduction
2 Pointwise Segmentation and a Theoretical Guarantee
3 Multipoint Segmentation
4 Conclusions
References
Multi-organ Segmentation Using Shape Model Guided Local Phase Analysis
1 Introduction
2 Method
2.1 Quadrature Filters and Model Guided Local Phase Analysis
2.2 Integrating Region-Based and Edge-Based Energy in the Level-Set Method
2.3 Hierarchical Segmentation Pipeline and Multi-scale Phase Analysis
3 Experiments and Results
4 Discussion and Conclusion
References
Filling Large Discontinuities in 3D Vascular Networks Using Skeleton- and Intensity-Based Information
1 Introduction
2 Methodology
2.1 Method Overview
2.2 Gap Filling
2.3 Generating the Second-Order Tensor Field T
2.4 Deriving a Saliency Map S and Preferential Directions D from T
2.5 Generating the Enhancement Map E
3 Results
3.1 Synthetic Data
3.2 3D Images of Tumour Vasculature
4 Discussion
References
A Continuous Flow-Maximisation Approach to Connectivity-Driven Cortical Parcellation
1 Introduction
2 Methodology
2.1 Iterative Markov Random Field Formulation
2.2 Continuous Max-Flow Optimisation
3 Results
4 Discussion
References
A 3D Fractal-Based Approach towards Understanding Changes in the Infarcted Heart Microvasculature
1 Introduction
2 Methods
2.1 Data Acquisition and Pre-processing
2.2 Segmentation
2.3 Fractal-Based Methods
3 Results
4 Conclusions
References
Segmenting the Uterus in Monocular Laparoscopic Images without Manual Input
1 Introduction and Background
2 Methodology
3 Experimental Results
4 Conclusion
References
Progressive Label Fusion Framework for Multi-atlas Segmentation by Dictionary Evolution
1 Introduction
2 Proposed Method
2.1 Dictionary Construction
2.2 Multi-layer Label Fusion
3 Experiments
3.1 Dataset and Parameters
3.2 Hippocampus Segmentation Results
4 Conclusion
References
Multi-atlas Based Segmentation Editing with Interaction-Guided Constraints
1 Introduction
2 Multi-atlas Based Editing Method
2.1 Extraction of Local Interaction Combinations
2.2 Selection of Training Labels with respect to User Interactions
2.3 Label Fusion Based on User Interactions
3 Experimental Result
4 Conclusion
References
Quantitative Image Analysis II: Microscopy, Fluorescence and Histological Imagery
Improving Convenience and Reliability of 5-ALA-Induced Fluorescent Imaging for Brain Tumor Surgery
1 Introduction
2 Concept of the Image Acquisition System
3 Imaging Formula
4 Experiment
4.1 Prototype System Building
4.2 Real Time Image Processing and Display
4.3 Quantitative Imaging Technique
5 Discussion and Conclusions
References
Analysis of High-throughput Microscopy Videos: Catching Up with Cell Dynamics
1 Introduction
2 Methods
2.1 Problem Formulation
2.2 Time Series Analysis
2.3 Dynamic Shape Model
2.4 MAP Segmentation and Association
3 Experimental Results
4 Summary and Conclusions
References
Neutrophils Identification by Deep Learning and Voronoi Diagram of Clusters
1 Introduction
2 Modeling Individual Cell Appearances by CNN
3 Modeling Cell Context by VDC
4 Experiments and Evaluation
5 Conclusions
References
U-Net: Convolutional Networks for Biomedical Image Segmentation
1 Introduction
2 Network Architecture
3 Training
3.1 Data Augmentation
4 Experiments
5 Conclusion
References
Co-restoring Multimodal Microscopy Images
1 Introduction
2 Data Acquisition
3 Methodology
3.1 Theoretical Foundation of Microscopy Image Restoration
3.2 Multimodal Microscopy Image Restoration Algorithm
3.3 Build a Look-Up Table for Better Initialization
3.4 Cell Segmentation and Classification Based on Co-restoration
4 Experimental Results
4.1 Qualitative Evaluation
4.2 Quantitative Evaluation
5 Conclusion
References
A 3D Primary Vessel Reconstruction Framework with Serial Microscopy Images
1 Introduction
2 Methods for 3D Vessel Reconstruction
2.1 Automated 2D Vessel Segmentation
2.2 Two-Stage Vessel Association with Vessel Cross-Sections
3 Experimental Results and Validation
4 Conclusion
References
Adaptive Co-occurrence Differential Texton Space for HEp-2 Cells Classification
1 Introduction
2 Method
2.1 Co-occurrence Differential Texton
2.2 HEp-2 Cell Image Representation in the Adaptive CoDT Feature Space
3 Experiments and Comparisons
3.1 Datasets
3.2 Experimental Results
4 Conclusion
References
Learning to Segment: Training Hierarchical Segmentation under a Topological Loss
1 Introduction
2 Method
3 Experiments and Results
4 Discussion and Conclusions
References
You Should Use Regression to Detect Cells
1 Introduction
2 Learning to Localize Cells
2.1 Defining the Proximity Score Map
2.2 Training and Evaluating a Regression Model
2.3 Detecting the Cells from the Proximity Score Map
3 Experimental Results
3.1 Datasets
3.2 Model Evaluation
4 Conclusion
References
A Hybrid Approach for Segmentation and Tracking of Myxococcus Xanthus Swarms
1 Introduction
2 Methodology
2.1 Context Builder
2.2 Cell Tracker
3 Experiments and Evaluations
4 Conclusions
References
Fast Background Removal in 3D Fluorescence Microscopy Images Using One-Class Learning
1 Introduction
2 Method
2.1 Estimating Background Noise
2.2 Detecting Undersized Windows
2.3 Recomputing Background Noise in "Undersized Windows"
3 Experiments and Results
4 Conclusions
References
Motion Representation of Ciliated Cell Images with Contour-Alignment for Automated CBF Estimation
1 Introduction
2 Ciliary Beating Frequency Estimation
2.1 Data Acquisition
2.2 Region Division
2.3 Cell Registration
2.4 Ciliary Beating Signal Extraction
2.5 Frequency Computation
3 Experimental Results
4 Conclusions
References
Multimodal Dictionary Learning and Joint Sparse Representation for HEp-2 Cell Classification
1 Introduction
2 Method
2.1 Supervised Dictionary Learning
3 Experiments and Results
3.1 Implementation Details
4 Conclusion
References
Cell Event Detection in Phase-Contrast Microscopy Sequences from Few Annotations
1 Introduction
-
-
-
2 Background Subtraction
3 Sparse Gaussian Processes
X
x1,
xN
xn
y
x|
Z
u
KZX
Z
X,
f
u|Z,X),
u
u|Z,X, y,
u|KZZQ
?
KZZQ
Q
KZZ
KZX(?
kT
K
kZx.
kZx
Z
x.
y|X,Z,
y|0,KT
K
KZX
?
Z
t|
f)
t
u|Z,X, y,
u|
X,
Z,
xn,
t)
X, u,Z)
u|X,Z, t)
u,
xn)
xn)
4 Collaborative Multi-output Gaussian Processes for Novelty Detection
X
Y
gj's
hi's
xn,
yn
xn)
xn),
xn)]
xn)
xn)
yn
y
yn,
I),
y
xn), Sn)
Sn]
xn)
y
I|
xn)
S
I)
xn)
yn)
xn)
yn)
Sn
Sn
xn)
xn)
4.1 Proposal Generation by Sampling
xn)
xn)
xj)
xn
xn)
5 Workflows
CGP-KL:
xn)
CGP-MSE:
xn)
OC-SVM:
150fps:
CGP-MSE
- MES:
xn)
Xu, yu)
Xu
yu
Xu, yu)
xn
- BALD:
xn)
Xu, yu]
Xu, yu, xn,
Xu, yu)
6 Experiments
x, x
x
x
6.1 Event Detection from Few Annotations
7 Discussion
References
Robust Muscle Cell Quantification
Using Structured Edge Detection
and Hierarchical Segmentation
1 Introduction
2 Structured Edge Detection
Random Decision Forest.
Structured Edge Detection.
3 Hierarchical Image Segmentation
x
x
x,
Muscle Cell Likelihood Generation.
4 Experiments
Experimental Results for Structured Edge Detection.
Experimental Results for Hierarchical Segmentation.
5 Conclusion
References
Fast Cell Segmentation Using Scalable Sparse
Manifold Learning and Affine
Transform-Approximated Active Contour
1 Introduction
2 Methodology
2.1 Scalable Sparse Manifold Learning for Shape Prior Modeling
Shape Prior Modeling via Manifold Learning:
1T
1T
Shape Inference and Cluster Update:
1Tai
1Tai
2.2 Fast Active Contour Model
Local Repulsive Deformable Model:
v)
v)
ni[-
ni
Affine Transform Approximation:
3 Experiments
4 Conclusion
References
Restoring the Invisible Details in Differential
Interference Contrast Microscopy Images
1 Introduction
1.1 Related Work
1.2 Our Proposal and Algorithm Overview
2 Methodology
2.1 Gradient Images
2.2 Bandpass Filter
2.3 Motion Magnification
2.4 Combine Forward and Backward Motion Images
3 Experimental Results
3.1 Qualitative Evaluation
3.2 Quantitative Evaluation
4 Conclusion
References
A Novel Cell Detection Method Using Deep
Convolutional Neural Network and
Maximum-Weight Independent Set
1 Introduction
2 Methodology
2.1 Cell Detection Using MWIS
x
x
x
x,
xTAx
x
x
x)
Step 1:
x)
x)
Step 2:
2.2 Deep Convolutional Neural Network
DCNN Architecture:
Foveation:
DCNN Training:
3 Experiments
4 Conclusion
References
Beyond Classification: Structured Regression
for Robust Cell Detection Using Convolutional
Neural Network
1 Introduction
2 Methodology
CNN-Based Structured Regression.
I
CNN Architecture.
Structured Prediction Fusion and Cell Localization.
Speed Up.
3 Experimental Results
Data Set and Implementation Details.
Model Evaluation.
Comparison with Other Works:
Ed
En
Ed
3.1 Conclusion
References
Joint Kernel-Based Supervised Hashing for Scalable Histopathological Image Analysis
1 Introduction
2 Methodology
Joint Kernel-Based Hashing:
Supervised Optimization:
0,
1Tµ
0
11T
1Tµ
0
3 Experiments
Experimental Settings:
Results and Analysis:
4 Conclusion
References
Deep Voting: A Robust Approach
Toward Nucleus Localization in Microscopy
Images
1 Introduction
2 Methodology
2.1 Learning The Deep Voting Model
Preprocessing:
vij
vij
vij
vij
vij
vij
Hybrid Non-linear Transformation:
non-linear transformation
Inference In Deep Voting Model:
vij
2.2 Weighted Voting Density Estimation
2.3 Efficient Nuclei Localization
y)
y,
3 Experiments
Data and Implementation Details:
Evaluate the Deep Voting Model:
Ed)
En)
DV-1),
DV-3),
PC). CNN+PC
DV-1.
IRV)
NERS)
ITCN)
Experimental Results:
DV-1
DV-3
CNN-PC
DV-3
DV-1, DV-3,
DV-1
DV-3
3.1 Conclusion
hybrid non-linear acti-vation
function.
References
Robust Cell Detection and Segmentation
in Histopathological Images Using Sparse
Reconstruction and Stacked Denoising
Autoencoders
1 Introduction
2 Methodology
2.1 Detection via Sparse Reconstruction with Trivial Templates
Adaptive Dictionary Learning.
vTvj)/(
vi
vj
vi
Probability Map Generation via Sparse Reconstruction with Trivial Templates:
pij
B
pij
Bc+e
B I][c e]
e
Im×m
pij
~B~c
d
c
e
1T c
~B
B I],~c
c e]
d
pij
B~~c)
2.2 Cell Segmentation via Stacked Denoising Autoencoders
X~
x) ~
W~x+b),
y
W, b}
W
y)
W
y
b
W
b
Z
X.~
z
x
x
x)~
WT
3 Experimental Results
Data set:
Detection Performance Analysis:
Segmentation Performance Analysis:
4 Conclusion
References
Automatic in Vivo Cell Detection in MRI
1 Introduction
2 Proposed Approach
2.1 Spot Modeling
S
2.2 Model Instantiation via Superpixel
S
S
2.3 Superferns Feature Extraction
S
G(
oi
G(
oi)
G(
G(
oi)
G(
G(
oi
G(
G(
2.4 Partition-Based Bayesian Classification
S,
S
3 MRI Database for Cell Detection
MRI:
MRI:
Labeling:
si
si
4 Experimental Results
Experimental Setup:
^s,
s ^
si
Performance and Comparison:
Superferns vs. Ferns:
oi
oi,
Diversity Analysis:
5 Conclusions and Future Work
References
Automatic Vessel Segmentation
from Pulsatile Radial Distension
1 Introduction
2 Methods
x
x
The Monogenic Framework:
u
u
u.
u
u|
u
u|
hj,
x
u
u)]
x
u
u)]
x
u
u)] hj,
x,
x
x
u)
x)]
Extraction of Local Motion from the Local Phase Vector:
r
hj,rj
x
hj,
hj,
h
hj+
hdiff,
rdiff
dN
rj(x
dN,
rdiff
x
dN,
dN,
dj(x)
x)
x
dj(x)).
Modeling of Radial Distension:
d(x,
x
d(x,
x,
d^
x,
d(x,
x,
d^
x,
x)
d^
x,
3 Results and Discussion
Materials and Experimental Setup:
Phantom Experiment:
dGT
d-dGT
Real Data Experiment:
Discussions:
4 Conclusions
References
A Sparse Bayesian Learning Algorithm
for Longitudinal Image Data
1 Introduction
2 Theory
0,P
P
2.1 Training Phase
X
X
Xn,
Xn
Xn,
Xn,
XTn
?n),
?n
IMn
1Mn1T
1T
IMn
1Mn
?n
X
0,C),
C
?
XTP
?
IN
?n
C
C|,
XT?
S
P
SXT?
P
2.2 Testing Phase
XN+
XN+
XN+
S
?
XN+
XT
?
1T
S
XN+
Xunknown
S
Xknown
S
2.3 Implementation
XT(XXT)-
P
3 Empirical Results
S0)
S1)
S2)
S1
S2,
S1
S2,
S2,
4 Discussion and Conclusion
References
Descriptive and Intuitive Population-Based
Cardiac Motion Analysis via Sparsity
Constrained Tensor Decomposition
1 Introduction
-
-
-
-
-
2 Methods
2.1. Re-orientation of Affine Matrices to Local Coordinates:
2.2. Extraction of Rotation, Stretch, Shear, and Translation:
2.3. 4-Way Tensor Decomposition:
2.4. Tucker Decomposition with Sparsity Constraints:
2.5. Combined Basis Computation of Multiple Groups:
3 Experiments and Results
0.697
0.847
4 Conclusion
Acknowledgements.
References
Liver Motion Estimation via Locally Adaptive
Over-Segmentation Regularization
1 Introduction
2 Methodology
Deformable Image Registration.
Filtering of the Deformation Field.
Guidance for Accurate Modeling of Complex Liver Motion.
SIG
SIG
U
Physiological Plausibility.
3 Evaluation and Results
Results on 4DCT.
Results on DCE-MRI.
4 Discussion and Conclusions
Acknowledgments.
References
Motion-Corrected, Super-Resolution Reconstruction for High-Resolution 3D Cardiac Cine MRI
1 Introduction
2 Image Reconstruction
2.1 Motion-Compensated 3D Cine Reconstructions
2.2 Super-Resolution Reconstruction
3 MRI Experiments
4 Validation
5 Results
6 Discussion and Conclusion
Acknowledgement.
References
Motion Estimation of Common Carotid Artery
Wall Using a H8 Filter Based Block Matching
Method
1 Introduction
2 Methodology
Prediction Step.
xn+
xn
yn
xn
vn,
zn
xn,
wn
vn
xn
yn
xn+
zn
zn,
xn+
Kn
Pn[I
RPn]
Pn+
Pn[I
R
Pn]
Qn,
xn+
xn
Kn(yn
xn),
Qn
Rn
wn
vn
Sn
Pn
I
Qn
Rn
Sn
I,
x1
Pn+
Pn
Kn+
Kn.
Updating Step.
Two Implementations.
xn
cn
xn
xn.
yn
c
cn+
xn
yn
xn
xn
yn
xn
yn
3 Results
4 Discussion and Conclusion
Acknowledgements.
References
Gated-tracking: Estimation of Respiratory
Motion with Confidence
1 Introduction
2 Materials
dj(
3 Method
3.1 Tracking Algorithm
3.2 Confidence Measure
dj(
c)
c
c(
c))
c)
c)
c)||.
c)
If
and d
d
do
else do
3.3 Evaluation
4 Results
5 Conclusion
References
Solving Logistic Regression with Group
Cardinality Constraints for Time Series Analysis
1 Introduction
2 Solving Sparse Group Logistic Regression
Model:
Algorithm:
3 Characterizing TOF Based on Cine MRIs
Extract Image Feature:
Measure Accuracy:
Grp.
Alternative Models:
No-Grp)
Rlx-Grp)
Classification Results:
4 Conclusion
Acknowledgement.
References
Spatiotemporal Parsing of Motor Kinematics
for Assessing Stroke Recovery
1 Introduction
2 Approach
2.1 Creating Candidate Foreground Regions
2.2 Robust Exemplar Classification
Max-projected Randomized Exemplar Classifiers:
Dictionary of Paw Classifiers:
Non-parametric Representation of Hand Posture:
2.3 Spatiotemporal Parsing
1
1(
1
2.4 Automatic Grasping Diagnostics
3 Experimental Results
Recording Setup:
Grasping Diagnostics:
4 Conclusion
References
Longitudinal Analysis of Pre-term Neonatal
Brain Ventricle in Ultrasound Images
Based on Convex Optimization
1 Introduction
Contributions:
2 Method
Optimization Model:
Sequential Convex and Dual Optimization:
Dual Optimization Formulation.
Proposition 1.
3 Experiments and Results
Image Acquistion:
Evaluation Metrics:
Results:
4 Discussion and Conclusion
Acknowledgments.
References
Multi-GPU Reconstruction of Dynamic
Compressed Sensing MRI
1 Introduction
2 Method
2.1 Compressed Sensing Formulation for DCE-MRI
Algorithm 1.
2.2 Multi-GPU Implementation
3 Result
3.1 Image Quality and Running Time Evaluation
3.2 Multi-GPU Performance Evaluation
4 Conclusion
Acknowledgements.
References
Prospective Identification of CRT Super
Responders Using a Motion Atlas and Random
Projection Ensemble Learning
1 Introduction
2 Materials
cine MR:
T-MR:
3 Methods
3.1 Spatio-Temporal Motion Atlas
Estimation of LV Geometry.
Estimation of LV Motion.
Spatial Normalisation.
Motion Reorientation.
vn
un
un
u
vn
vatlas
up))
vn
AHA Segmentation.
vatlas
vatlas
3.2 Random Projection Ensemble Learning
vatlas
vatlas
vn
vn
X
v1
vN]
X
Y
X
Rl
Z
R
X, Z
Z
Y.
R
Rl
X
xtest
RF
Rl
4 Experiments and Results
X
xtest.
X16, X6
X3
X16
X6
X3
5 Discussion
Acknowledgements.
References
Motion Compensated Abdominal Diffusion
Weighted MRI by Simultaneous Image
Registration and Model Estimation (SIR-ME)
1 Introduction
2 Methods
2.1 Simultaneous Image Registration and Model Estimation
(SIR-ME) for Motion Compensated DW-MRI Parameter
Estimation
2.2 Optimization Scheme
Signal Decay Model (
Estimation:
Estimation of Transformation
Reconstruction of High SNR DW-MRI Signal
Estimation of Transformation
3 Results
4 Conclusions
References
Fast Reconstruction of Accelerated Dynamic
MRI Using Manifold Kernel Regression
1 Introduction
2 Methods
2.1 Manifold Learning
x)
xi
xj)
xi
xj)
xi
xj
L
D-W.
x)
xi
xj)
xi
xj)
Lx
xTDx
x
Lx
W
ys
yt
y
y
y
2.2 Manifold Learning of Under-Sampled k-space
2.3 k-space Interpolation Using Manifold Kernel Regression
yf
xf
xi)yi yf
xf
xj)
yi
xf
xf
x
x
xf
xi)
2.4 Algorithm Summary
3 Experiments and Results
3.1 Manifold Creation
3.2 Comparison to State-of-the-Art Compressed Sensing
Algorithms
4 Discussion
References
Predictive Modeling of Anatomy
with Genetic and Clinical Data
1 Introduction
2 PredictionModel
2.1 Subject-Specific Longitudinal Change
0,
2.2 Learning
2.3 Prediction
3 Model Instantiation for Anatomical Predictions
3.1 Anatomical Phenotype
3.2 Health Similarities
4 Experiments
4.1 Volumetric Predictions
4.2 Anatomical Prediction
5 Conclusions
Acknowledgements.
References
Joint Diagnosis and Conversion Time Prediction of Progressive Mild Cognitive Impairment (pMCI) Using Low-Rank Subspace Clustering and Matrix Completion
1 Introduction
2 Materials and Preprocessing
2.1 Materials
2.2 Preprocessing and Feature Extraction
3 Method
3.1 Notation
X
X
X
Y
M
M)
M
I
3.2 Low-Rank Matrix Completion (LRMC)
Xak+
X1]
ak
1
ak
X,
X1Y]
Z
MO
ZO
Z
ZOl
MOl
ZOs
MOs
QcZ
ZOl
MOl
ZOs
MOs
Qc
Qc
I.
QcZ
Z
Q(
Z
Qc
X
3.3 Incomplete Low-Rank Representation (ILRR)
X,
A
E
AX
E,
XO
X,
AX, A
E
A
E,
D
AX,
A,
3.4 Spectral Clustering
A
X
D).
A
A
A
A)
0.
L
S
S
A.
L
V
V
V
X(
D(
Y(
X(
1Y(
D(
1Y(
4 Results and Discussions
A
AX
5 Conclusion
Acknowledgment.
References
Learning with Heterogeneous Data
for Longitudinal Studies
1 Introduction
2 CO-MKL: Multiple Kernel Learning with Covariate
2.1 Background
2.2 Proposed Approach
3 Materials and Methods
4 Results
-
-
-
-
-
-
5 Conclusions
References
Parcellation of Infant Surface Atlas Using Developmental Trajectories of Multidimensional Cortical Attributes
1 Introduction
2 Method
2.1 Dataset and Cortical Surface Mapping
2.2 Atlas Parcellation Using Cortical Developmental Trajectories
3 Results
4 Conclusion
References
Graph-Based Motion-Driven Segmentation
of the Carotid Atherosclerotique Plaque
in 2D Ultrasound Sequences
1 Introduction
2 The Proposed Methodology
2.1 Group-Wise Image Registration
2.2 A Mutual-Information Map Reflecting Kinematic Dependencies
2.3 Graph-Based Segmentation of the Mutual-Information Map
g)
f
g,f
3 Evaluation Process and Results
4 Discussion
References
Cortical Surface-Based Construction of Individual Structural Network with Application to Early Brain Development Study
1 Introduction
Violation of Cortical Topological Properties.
Sensitivity to Intensity Heterogeneity.
Influence by Patch Size Heterogeneity.
2 Methods
Subjects and Image Acquisition.
Image Preprocessing.
Cortical Surface Reconstruction and Registration.
Construction of Individual Networks.
Network Metrics.
3 Results
4 Conclusion
References
NEOCIVET: Extraction of Cortical Surface and Analysis of Neonatal Gyrification Using a Modified CIVET Pipeline
1 Introduction
2 Methods
A)
B)
C)
D)
E)
F)
G)
H)
I)
J)
K)
L)
2.1 Subjects and MRI Acquisition
2.2 Classification of Brain Tissues
x),
x
x)
x)
x)
x)
x)
2.3 Surface Fitting
2.4 Surface-Based Morphological Measurements
2.5 Construction of Age-Specific Surface Templates and Registration
2.6 Other Technical Considerations and Parameter Selection
3 Application and Clinical Utility of NEOCIVET
3.1 Statistical Analyses
3.2 Results
4 Discussion and Conclusion
References
Learning-Based Shape Model Matching:
Training Accurate Models with Minimal Manual
Input
1 Introduction
2 Methods
2.1 Groupwise Non-rigid Registration
2.2 RF Regression-Voting in the CLM Framework
3 Experiments and Evaluation
Datasets:
3.1 Guided Groupwise Non-rigid Registration and Refinement
3.2 Estimation of GNR Accuracy in the Absence of Ground Truth
3.3 Performance of GNR-trained RFRV-CLM on New Images
4 Discussion and Conclusions
Acknowledgments.
References
Scale and Curvature Invariant Ridge Detector
for Tortuous and Fragmented Structures
1 Introduction
2 Methods
Curved-Support Gaussian Models.
Unsupervised SCIRD.
Supervised SCIRD.
3 Experiments and Results
Datasets.
IVCM
BF2D
VC6
Parameters Setting.
Results and Discussion.
4 Conclusion
References
Boosting Hand-Crafted Features
for Curvilinear Structure Segmentation
by Learning Context Filters
1 Introduction
2 Proposed Approach
Unsupervised Filter Learning.
D
x(
x(
x
V(S+
x(
X),
Dc(
c(
d
c(
x
d
D.
Learning Context Filters.
Description Vector and Supervised Classification.
L
D
In,
D
In
OOF(
L
L
3 Experiments
Datasets.
IVCM
BF2D
VC6
Experimental Setup.
Results and Discussion.
4 Conclusion
References
Kernel-Based Analysis of Functional Brain
Connectivity on Grassmann Manifold
1 Introduction
2 Functional Connectivity Classification on Manifold
2.1 Grassmann Manifold
2.2 Identification of Discriminative Functional Connections
3 Material
Algorithm 1.
4 Results and Discussion
Grass-Kernel)
GK-LogE)
GK)
LK)
5 Conclusion
References
A Steering Engine: Learning 3-D Anatomy Orientation Using Regression Forests
1 Introduction
2 Methods
2.1 Problem Definition
2.2 Regression-Based Anatomy Colatitude Prediction
2.3 Iterative Colatitude Prediction and Patch Reorientation
2.4 Rotation-Invariant Integral Image for Fast Feature Calculation
3 Experimental Results
Local Anatomy Orientation Detection:
Vertebrae
Aorta
Anatomical Structure Tracing:
4 Conclusions
References
Automated Localization of Fetal Organs in MRI
Using Random Forests with Steerable Features
1 Introduction
2 Method
3 Experiments
4 Conclusion
References
A Statistical Model for Smooth Shapes in Kendall Shape Space
1 Introduction and Related Work
2 Methods
2.1 Statistical Shape Model Using Riemannian Distances and Smoothness Priors
2.2 Efficient Model of Likelihood of Image Data Given Object Shape
2.3 Parameter Inference Using ExpectationMaximization
x ¯
¯x|
z ¯
¯
¯x,Z
x,¯¯z|
z ¯
x,¯¯zs|
zs ¯
¯Z
¯x,
¯x
¯zs}
Update Mean
¯zs}
Update Covariance
Update Similarity Transforms:
¯zs}
2.4 Sampling Smooth Shapes in Kendall Shape Space
3 Results
Validation on Simulated Data:
Evaluation on Clinical Data:
Conclusions.
References
A Cross Saliency Approach to
Asymmetry-Based Tumor Detection
1 Introduction
2 Cross Saliency
2.1 Cross Patch Distinctness
2.2 Patch Flow
2.3 Combining the Measures
3 Experiments
3.1 Unsupervised Tumor Detection on BRATS2014
3.2 Tumor Detection in Breast Mammogram
3.3 Stability under Variation in Mid-Sagittal Plane Estimation
4 Conclusion
References
Grey Matter Sublayer Thickness Estimation
in the Mouse Cerebellum
1 Introduction
2 Methods
2.1 Cerebellum Extraction and Tissue Segmentation
2.2 Fissure Extraction
2.3 Grey Matter Thickness Estimation
2.4 Purkinje Layer Extraction and Sublayer Thickness Estimation
2.5 Grey Matter Function Aware Region Parcellation
3 Experimental Data and Validation
4 Conclusion and Discussion
References
Unregistered Multiview Mammogram Analysis
with Pre-trained Deep Learning Models
1 Introduction
Literature Review:
Contributions:
2 Methodology
x(
c(
m(
y(
x
xCC, xMLO
xCC, xMLO
c
cCC, cMLO
cCC, cMLO
m
mCC,mMLO
mCC,mMLO
y
y
y
y
y
xx
y
Convolutional Neural Network:
xk(
xk-
Wk(
xk(
xk-
xL
y(
y
x
y
xMLO
y
xCC
y
cMLO
y
cCC
y
mMLO
y
mCC
xMLO,
y
xMLO,
xCC,
cMLO,
cCC,
mMLO,
mMLO,
3 Materials and Methods
4 Results
5 Discussion and Conclusion
References
Automatic Craniomaxillofacial Landmark
Digitization via Segmentation-Guided
Partially-Joint Regression Forest Model
1 Introduction
2 Automatic CMF Landmark Digitization
2.1 Traditional Regression Forest Model for Landmark Digitization
2.2 Segmentation-Guided Partially-Joint Regression Forest
(S-PRF) Model
Partially-Joint Regression Forest Model:
Segmentation-Guided Strategy:
d
3 Experiments, Results and Discussion
3.1 Data Description and Parameter Setup
3.2 Experimental Results
Effect on the Segmentation-Guided Strategy:
Effect on the Use of Partially-Joint Versus Fully-Joint Models:
Effect on Adding Additional MSCT Images into the Training Dataset:
Effect on Features:
Comparison with Registration-Based Landmark Detection:
Effect on the Use of Scaling Coefficient:
4 Conclusion
References
Automatic Feature Learning for Glaucoma
Detection Based on Deep Learning
1 Introduction
2 Method
2.1 Feature Learning Based on Deep Convolutional Neural Network
Convolutional Layers.
L
L
L
L
L
L
L
w
1
w
Multilayer Perceptron Convolution Layers.
Contextualizing Training Strategy.
2.2 Glaucoma Classification
Disc Segmentation.
Dropout and Data Augmentation.
Automatic Classification by Softmax Regression.
3 Experiments
3.1 Evaluation Criteria
3.2 Experimental Setup
3.3 Comparison of Different Types of CNN Architectures
-
-
-
-
-
3.4 Glaucoma Diagnosis
4 Conclusion
References
Fast Automatic Vertebrae Detection and
Localization in Pathological CT Scans - A Deep
Learning Approach
1 Introduction
2 Methods
2.1 Point Selection
2.2 Feature Extraction
2.3 Deep Neural Network
2.4 Centroid Estimation
2.5 Refinement
3 Experiments and Results
4 Discussion
Acknowledgement.
References
Guided Random Forests for Identification of Key Fetal Anatomy and Image Categorization in Ultrasound Scans
1 Introduction
2 Guided Random Forests
2.1 Localizing Object of Interests
2.2 Feature Sets for e the Learning Algorithm
2.3 Training and Testing
3 Experiments
3.1 Dataset
3.2 Evaluation Protocol and Metrics
4 Results
5 Discussion and Future Work
References
Dempster-Shafer Theory Based Feature
Selection with Sparse Constraint for Outcome
Prediction in Cancer Therapy
1 Introduction
2 Backgrounds on Dempster-Shafer Theory
3 Method
R
4 Experimental Results
5 Conclusion
Acknowledgements.
References
Disjunctive Normal Shape and Appearance
Priors with Applications to Image Segmentation
1 Introduction
2 Disjunctive Normal Shape Model
RD
x
RD
x)
x
RD
x)},
x)
x)
x)
x)
x)
x)
x).
x)
x)).
x)
x
x
x)
x)
3 DNSM Shape Priors
W
Wt)
x)
x))
x)
x)
x)
x)
Wt.
Wi,
Wi
Wi)
Wi,Wi
Wi,Wi
Wijk Wt
Wi,Wi
WWt
Wijk
W
Wi,Wi
Wi)
W,Wt).
4 DNSM Appearance Priors
W),
W)
x)
x))
x)
x)
x)
x))
x))
x)
x)) xk
5 Segmentation Algorithm
6 Experiments
Prostate Central Gland Segmentation:
Full Prostate Segmentation:
7 Conclusion
Acknowledgments.
References
Positive Delta Detection for Alpha Shape Segmentation of 3D Ultrasound Images of Pathologic Kidneys
1 Introduction
2 Methods
2.1 Local Phase-Based 3D Positive Delta Detection
2.2 Patient Specific Pos sitional Maps via Alpha Shapes
2.3 Active Contours Formulation
2.4 Complete Segmentation Framework of Renal Structures in 3DUS
3 Results and Discussion
4 Conclusions
References
Non-local Atlas-guided Multi-channel Forest
Learning for Human Brain Labeling
1 Introduction
2 Method
Notations.
A
A.
2.1 Atlas-guided Multi-channel Forest Learning
A,
Sampling Strategy:
Feature Extraction:
1 channel
channels
2.2 Single-ROI and Multi-ROI Labeling
Single-ROI Labeling:
1)
2)
3)
Multi-ROI Labeling:
2.3 Haar-Based Multi-Class Contexture Model (HMCCM)
Training:
Testing:
3 Experimental Results
Parameters:
LONI-LPBA40 Dataset:
IXI Dataset:
4 Conclusion
References
Model Criticism
for Regression on the Grassmannian
1 Introduction
Contributions.
2 Model Criticism for Regression in Euclidean Space
Review of Kernel-Based Two-Sample Testing [3].
Model Criticism Using Two-Sample Testing.
3 Model Criticism for Regression on the Grassmannian
Drawing Random Samples on the Grassmannian.
Yobs
Y
Yobs
Y
Y
Y
Zi
Y
Y
Is
Y
Y
Zi
Is
Yest
Yest
Y
Y
Zi
Y
Yest
Kernels for Model Criticism on the Grassmannian.
X,Y)
X
Y
Yi)
X),
Y))
X
Y
Model Criticism on the Grassmannian.
Y
Yobs
Yest
Yest
Yobs
Ya
Yb
Ya
Yb
4 Experimental Results
Synthetic Data.
(1) Different Data Distributions.
(2) Different Regression Models.
Real Data.
(1) Corpus Callosum Shapes.
(2) Rat Calvarium Landmarks.
5 Discussion
Acknowledgments.
References
Sample Size Estimation for Outlier Detection
1 Introduction
2 Methods
2.1 A Confidence Interval for
2.2 Estimating Sample Sizes
3 Experiments
3.1 Plausibility Test: Our Formula vs. Monte Carlo Estimate
3.2 Sample Size Estimation for Given
3.3 Application to Real Data
4 Conclusion
Acknowledgement.
References
Multi-scale Heat Kernel Based Volumetric
Morphology Signature
1 Introduction
2 Methods
2.1 Theoretical Background
2.2 Discrete Multi-scale Volumetric Morphology Signature
Ti
T)(Ti
T)
Ti
T
Ti
Ti
T
2.3 Internal Structure Feature Selection
3 Experimental Results
3.1 Synthetic Data Results
3.2 Application to Alzheimer's Disease
3.3 Comparison with Freesurfer Thickness Feature
4 Conclusions and Future Work
References
Structural Brain Mapping
1 Introduction
Brain Net.
Approach.
2 Theoretic Background
Graph Embedding.
Harmonic Map.
g)
g
3 Computational Algorithms
Step 1: Isomorphic Graph Embedding.
Step 2: ConstrainedHarmonicMapping.
Algorithm 1.
Input:
4 Experiments
Brain Net Extraction.
Structural Brain Mapping.
5 Discussion
6 Conclusion
Acknowledgments.
References
Author Index

System requirements

Save as PDF Copy link into clipboard

Schweitzer Fachinformationen

Medical Image Computing and Computer-Assisted Intervention - MICCAI 2015

Description

More details

Other editions

Additional editions

Content

System requirements