27.12.2012 Views

ARUP; ISBN: 978-0-9562121-5-3 - CMBBE 2012 - Cardiff University

ARUP; ISBN: 978-0-9562121-5-3 - CMBBE 2012 - Cardiff University

ARUP; ISBN: 978-0-9562121-5-3 - CMBBE 2012 - Cardiff University

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

ehavior. Other researchers have measured the mechanical response of in-vivo liver<br />

tissue [9-11]. However, they do it by means of invasive or open surgery.<br />

In this situation, the use of Medical Image Techniques can play an important role to<br />

estimate the parameters of biomechanical models since they can be obtained through<br />

non-invasive methods avoiding surgery. Some researchers have used Elasticity Imaging<br />

Techniques as Magnetic Resonance Elastography (MRE) [12-13]. MRE uses<br />

propagating acoustic shear waves to measure the stiffness of the soft tissue.<br />

Nevertheless MRE only allows the estimation of the Young’s modulus and viscous<br />

effects, so parameters of non-linear elastic models cannot be estimated. Furthermore,<br />

this technique requires additional and complex hardware to measure and create the<br />

mechanical wave.<br />

Therefore, the similarity coefficients commonly used for segmentation validation in<br />

Medical Image can be quite suitable used over liver volumes reconstructed from<br />

Computer Tomography (CT). They can be used to estimate how accurate are the<br />

parameters comparing a real deformed liver volume with the liver volume of the<br />

simulated deformation. This method allows estimating the parameter of any model and<br />

only using a CT Machine, that is, no additional hardware is needed in the Hospital.<br />

Inside this framework, this paper takes the last approach to estimate the parameters of<br />

several biomechanical models. An Iterative Search Algorithm (ISA) is used to find the<br />

optimum parameter of each model where the results of the Geometric Similarity<br />

Function (GSF), based on similarity coefficients, provides information to the ISA about<br />

the direction of the search.<br />

3. MATERIALS AND METHODS<br />

This section is divided in two subsections. The first one explains everything related to<br />

obtain the model geometry of the liver and the simulation of its deformation due to the<br />

patient breathing using biomechanical models. The second subsection explains the<br />

method used to estimate the models parameters based on a Geometric Similarity<br />

Function (GSF) and an iterative search algorithm (ISA).<br />

3.1. Simulation of the human liver deformation due to the breathing.<br />

The biomechanical models used for the simulations were two:<br />

• Linear elastic model: Two parameters characterize the behavior of the material,<br />

Young’s modulus E and Poisson’s ratio υ. These two parameters were taken<br />

from the literature for the reference deformation [1], which would be equivalent<br />

to the real deformation (E=7800 Pa and υ=0.3).<br />

̅ ̅<br />

• Mooney-Rivlin’s model: The strain energy potential for this model with two<br />

parameters is defined as Eq. 1 indicates:<br />

= − 3 + − 3 +<br />

2<br />

− 1<br />

(1)<br />

Where C10 and C01 are material constants, ̅ and ̅ are the first and second<br />

deviatoric strain invariant respectively, K0 is the Bulk modulus and J is the<br />

determinant of the elastic deformation gradient. For the reference deformation

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!