Biomedical Engineering Reference
In-Depth Information
(1) Building and visualizing 3D patient model from medical images automatically
(2) Building computational biomechanics models for simulations of deforming hard
and soft musculoskeletal tissues
(3) Building prediction and diagnosis models for therapy
(4) Developing integrated patient models for medical education
(5) Validation and evaluation methodologies for 3D patient modeling.
1.3 Building and Visualizing 3D Patient Model from Medical
Images Automatically
1.3.1 Multi-Scale Biomechanical Modeling
Intensive studies have been carried out on constitutive modeling of articular cartilage
[
1
-
4
]. The existing theories of cartilage tissue and the experimental data avail-
able in literature [
5
-
7
] allow the development of a complex, but accurate, bipha-
sic model of the tissue behavior [
8
]. We should ideally consider the constitutive
modeling of degenerated cartilage [
1
,
9
], and cartilage growth [
10
,
11
], the tissue
differentiation models based on biphasic mechano-regulation theory [
12
] extended
to model tissue differentiation during osteochondral defect repair [
13
], cell migra-
tion/proliferationmodels using randomwalkmodeling [
1
] as well as developments of
earlier diffusion based approaches [
10
], bone remodeling algorithms based on com-
bined strain/damage-adaptive theory, and single cell combined tensegrity-continuum
finite element models. Figure
1.4
shows an example of tissue modeling.
To alleviate the complexity of conducting a simulation involving all levels, we
retain to model the musculo-skeletal system at a higher abstraction level with ide-
alized joints. This choice keeps the inverse dynamics computation efficient, hence
allowing to exploit it within a decoupled simulation framework where it provides a
global view of the individual muscle forces acting at the joint level. This information
is then exploited to determine the contact distribution for the patient bone shape in
accordance with the tissue level simulation. Conversely, given a modified bone shape
the posture will be adjusted by constrained optimization to preserve the equilibrium
of the weight bearing experiments. To achieve this task, first the patient-specific
muscles action lines will be identified from the medial surfaces of the muscle 3D
meshes. Then, given the kinematic and reaction forces data of the weight bearing
experiments, Inverse Dynamic computations will be made at the level of the lower
limb musculoskeletal system. The resulting muscle activation and force distribution
will be used at the joint level simulation where the relative location and move-
ment of joint surfaces is investigated to determine the corresponding contact areas.
At that stage, the stress-strain distribution and visco-elastic biphasic behavior is
evaluated with the previous task simulation. We will study the joint system at two
levels. At the tissue level, the different mechanical properties of each joint compo-
nent are estimated or retrieved from the data acquisition and a strain/stress analy-
