Biomedical Engineering Reference
In-Depth Information
differences in bony geometry and muscle attachment sites. To understand the effects
of joint disease, a subject's unique articulation shape and cartilage thickness must
be known, both of which contribute to the contact pressure distribution. The need to
customize musculoskeletal anatomy is an essential step in the modeling process, if
the predictions of computer models are to be useful to clinicians [
44
].
Scheys et al. [
45
] have demonstrated the inaccuracy of gait kinematics calculated
from scaled generic models in subjects with increased femoral anteversion. Since
the results of simulations are often sensitive to the accuracy of the functional mus-
culoskeletal model, individualized musculoskeletal models may be a better alterna-
tive. Hence, for an accurate representation of subject-specific anatomical structures,
segmentation of medical images are extensively used. Medical imaging techniques
such as Magnetic Resonance Imaging (MRI), Computed Tomography (CT) and also
Positron Emission Tomography (PET) are used in order to generate 3D volumetric
data sets of the human body and to study in vivo the complex geometric relation-
ships among the muscles, bones, and other structures [
46
,
47
]. However, it is time
consuming and requires extensive imaging protocols to capture the muscle and joint
geometries at different limb positions. Subject-specific musculoskeletal modeling
also addresses the problem of image segmentation, which consists of extracting
anatomical structures from medical image data such as MRI. Semi-automatic or
fully automatic segmentation methods are fast but inaccurate since muscle distinc-
tion is often difficult or impossible to assess with currently used methods. Thus, soft
tissue volumetric representations are most often and most accurately acquired by
defining contours manually. Blemker et al. [
20
] built for example volumetric finite
element representations of muscles from manually segmented MRIs.
6.2.3.3 Preparing Anatomical Data for Simulation
The anatomical data resulting from artist models or segmented scans usually need to
be processed to produce a model that can be simulated. In this section we consider
two case studies.
In our first case study, we want to focus on the production of FEM-ready volu-
metric meshes from raw segmented surface models of medical images [
48
]. At first,
it comes handy to store areas such as attachment sites and tendon vertices in an
index-invariant structure by defining them by geometry as some tasks may change
the amount and ordering of the vertices. Then smoothing out the surface geometries
is usually performed by a three-parameter low-pass filter [
49
] to remove acquisi-
tion artifacts. Missing entities can be generated semi-automatically such as tendon
extremities but themost important step is the resolution of self-intersections and over-
laps. While using Boolean operators for solving overlaps, Peeters and Pronost [
48
]
proposed an algorithm to remove self-intersections in anatomical entities. The fi-
nal steps consist in generating the volume meshes with the relevant materials and
designing the FEM constraints from the index-invariant structure of the first step.
In our second case study, wewant to create a realistic and comprehensivemodel for
the human neck, starting from the surfacemodels of an anatomical artistic human data
