Texture-Based FEM Registration for Prostate MR Images

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional methods for registering pre-External Beam Radiation Treatment (EBRT) and post-EBRT MR images using Finite Element Models (FEMs) are limited by rigid initial registration and assume accurate node correspondence, leading to sub-optimal accuracy in assessing changes due to non-linear deformations and shrinkage in treated tissues, which complicates the evaluation of EBRT effectiveness.

Innovation Solution

A texture-based FEM registration scheme that constructs a FEM from the pre-EBRT image, captures surface and internal architecture, and maximizes mutual information between texture transformations of pre-EBRT and post-EBRT images using a biconjunctive gradient stabilized method and particle swarm optimization to determine displacements, thereby improving registration accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If conventional rigid registration methods are used for pre-EBRT and post-EBRT images, then the registration process is simple and fast, but the registration accuracy deteriorates due to non-linear deformations and shrinkage in treated tissues

Engineering Contradiction:

Improveregistration speedVSAvoidregistration accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The patent transitions from rigid registration (6 degrees of freedom) to elastic registration with FEM, changing the registration parameters to include tissue elasticity, compressibility, and non-linear deformation characteristics. This allows the model to adapt to EBRT-induced tissue changes while maintaining computational efficiency through optimized FEM solving algorithms.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent replaces conventional image processing-based registration with a physics-based FEM approach that uses material properties (elasticity, compressibility) to model tissue deformation. This substitution enables accurate representation of non-linear tissue changes while the use of stabilized solution methods and optimization algorithms maintains computational efficiency.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If FEM-based elastic registration is used to account for non-linear deformations, then registration accuracy improves, but computational cost increases

Engineering Contradiction:

Improveregistration accuracyVSAvoidcomputational cost

Core Design Contradiction:

Measurement precisionVSUse of energy by stationary object

Solution Approach 1:

The patent performs preliminary rigid registration to establish initial node correspondences between pre-EBRT and post-EBRT images before applying FEM-based elastic registration. This preliminary alignment reduces the computational burden of the subsequent elastic registration by providing a better starting point for the optimization algorithm.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent uses the image data itself to drive the FEM registration process through mutual information maximization, eliminating the need for external force measurements or manual boundary condition specification. The algorithm automatically determines optimal deformations based on image similarity, reducing computational overhead.

Inventive Principle:

Principle #25Self-service

3Device complexity

If conventional FEM registration assumes accurate node correspondence, then the registration process is simplified, but accuracy deteriorates when node correspondences are inaccurate due to significant tissue deformation

Engineering Contradiction:

Improveregistration process complexityVSAvoidnode correspondence accuracy

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The patent implements an iterative optimization process where the FEM registration is refined by maximizing mutual information between the deformed pre-EBRT image and the post-EBRT image. This feedback mechanism continuously adjusts node correspondences and material properties to improve alignment accuracy, overcoming the limitations of single-step conventional FEM methods.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The patent employs dynamic material properties for the FEM model, allowing elasticity and compressibility parameters to vary throughout the tissue volume rather than being uniform. This dynamic approach better represents actual tissue behavior under EBRT and improves the accuracy of deformation modeling and node correspondence.

Inventive Principle:

Principle #15Dynamics

Data Source

PatentUS9262583B2Image similarity-based finite element model registration

Publication Date: 2016.02.16 CASE WESTERN RESERVE UNIV

AI summary

Apparatus, methods, and other embodiments associated with evaluating global deformations and local deformations in a prostate are described. One example apparatus includes logics that evaluate global and local deformations in a prostate and register a pre-External Beam Radiation Treatment (EBRT) three dimensional (3D) magnetic resonance (MR) image with a post-EBRT 3D MR image. An image acquisition logic acquires a pre-EBRT image and a post EBRT image of an organ, item, or volume. An image texture information logic extracts image texture information from the pre-EBRT and post-EBRT images. A finite element model (FEM) transformation logic constructs a FEM of the volume imaged in the pre-EBRT image, deforms the FEM, deforms the pre-EBRT image as a function of the deformed FEM, and maximizes the image texture similarity between the deformed pre-EBRT image and the post-EBRT image. A registration logic registers the pre-EBRT image with the post-EBRT image based on the transformation.