What is image registration?
The process of transforming different sets of data into one coordinate system
What does image registration involve?
Estimating a smooth, continuous mapping between the points in one image and those in another
What can be determined from The parameters that encode the mapping?
The relative shapes of the images
What is the objective of image registration?
Determine the single “best” set of values for these parameters
What do many registration approaches use?
Small deformation model
What are the two broad categories of parameterisation?
- The small deformation framework does not necessilary preserve topology - although if the deformations are relatively small, then it may still be preserved
- The largest-deformation framework generates deformations that have a number of elegant mathematical properties such as enforcing the preservation of topology
What is the aim of image registration?
Estimate optimal geometric mapping between different images
Transform images so that they match in terms of:
- orientation (position)
- shape (anatomy)
Why is image registration necessary?
- Time series: motion correction
- different images are acquired at different time points
- motion: breathing, wiggle - Longitudinal:
- disease progression - in neurodegenerative diseases - look at inflammation and atrophy
- developmental changes
- plasticity changes
- clinical trials endpoints - Multi-modal
- diagnosis
- surgical planning - Template space
- group analyses
- replication
- atlas-based fusion
What is the basics of image transformation?
- When you get an image, read it into the matrix
- Have 8 by 8 matrix
- Each matrix/pixel in the matrix has a value
- white is 0
- black is 1 - Brain images are not binary, but intensity is a gradient
- Each value has an index- coordinates of the value of the matrix
What is translation?
Translations are movements in the x and y direction
Move forward, back, left right up and down
For rotation the origin is the same
For translation you shift the origin
B value is the transformation - you move by 2 points in the y direction
What is affine transformation?
It is a function mapping an affine space onto itself that preserves the dimension of any affine is spaces and also preserved the ratio of the lengths of parallel line segments
What are examples of types of transformation?
- Rigid: 6 DoF
- rotations (roll, pitch, yaw)
- translations (x,y,z) - Affine: 12 DoF
- 6 rigid
- scaling (width,height, depth)
- shearing (x,y,z)
What does affine transformation not necessarily preserve?
Angles between lines or distances between points, though it does preserve ratios of distances between points lying on a straight line
What is affine transformation?
It is the matrix you apply to an image in order to be able to transform it
The information that the matrix includes is the rotation and translation
What is the differences between a rigid and affine transformation?
- Rigid transformation is only 6 degrees of freedom - 3 rotations and 3 translations
- Affine transformation has 12 degrees of freedom - you have scaling and shearing
What is the principle of image registration?
- You have a target image and reference image
- You want to move the target image to the reference image
- Read it into a matrix
- Apply transformation matrix
What is image registration the process of?
- Estimating the transformation matrix that will match the target image to reference image
- Transforming (warping/reorienting) the target image so that it matches the reference image geometrically as close as possible
How do we match images?
- Manually
- overlay images
- visually guided realignment
- check alignment
- repeat until perfect match
But
- time consuming
- potential bias
- how do we handle the complex deformations?
- Read images in the same coordinate system
- estimate some transformation
- evaluate results of transformation
- repeat until you reach a point when there is no more improvement
What is the image registration framework?
- Deformation model
- Objective function
- Optimisation method
Now do you measure similarity between two images?
- Sum of square differences (SSD)
- Cross-correlation (CC)
- Mutual information (MI)
Affine transformation type
3 translation
3 rotations
3 scaling
3 shearing
What is diffeomorphic registration?
- Preserves topology
- maintains anatomical neighbourhood relationship and connectivity - Transformations are
- one to one
- smooth
- continuous
- invertible i.e. the compositions of the inverse and forward transformation results in approximately the identity transform - Flow fields (u) are estimated for each image to describe the transformation
What is the aim of feature-based registration?
Match geometrical features:
- landmarks
- surface patterns
Surface-based: better for cortical folds than volumetric
What are the different types of transformation?
- Rigid transformation
- Affine transformation
- Non-rigid transformation
- Landmark transformation
- Diffeomorphic transformation
-
Relaxation Theory & Measurement - T1 & T294
-
Image Registration36
-
Analysis of Quantitative MRI54
-
Physiology & Contrast Mechanisms for mapping brain activity58
-
Bolus MR/CT37
-
Magnetisation Transfer Theory, Measurement and Analysis60
-
Arterial Spin Labelling59
-
Experimental Design & Data Acquisition for Mapping Brain Activity (Block Design)93
-
The General Linear Model and Single-Subject & Group Analyses45
-
Resting State & Event-Related fMRI37
-
Voxel-Based Morphometry82
-
Principles of ROI volumetry26
-
An Overview of Medical Image Segmentation33
-
Connectivity: Dynamic Causal Modelling (DCM) for fMRI67
-
Diffusion-Weighted Imaging (DWI)89
-
Diffusion Tensor Imaging48
-
Analysis, Tractography and Applications in Diffusion Imaging47
-
Study Design: From group comparisons to individual predictions24
-
Multi-Modal Imaging - Motivation, Synchronicity, Instrumentation & Data Quality Issues34
-
Data Analysis - MRI/fMRI/DTI23
-
Fusion, Co-Registration, Technical Issues & Limitations32
