Contact Modeling and Collision Detection in Human Joints - 3D Multiscale Physiological Human - page 191

applications such as musculo-skeletal simulation as it is difficult to infer precisely the
relative range of motion from scanned or reconstructed 3D models. The estimated
range of motion was also an accurate value, which is another advantage comparing
to the collision detection based methods, where we have to define rotation steps as
a resolution. The results were also validated by confirming that the found range of
motion is never less than the full capsule's range of motion, and it is even almost the
same of full hip capsule's range of motion for flexion/extension [ 3 , 13 ].
8.5 Importance of Joint Center of Rotation in Medical
Simulations
In orthopedic simulations the behavior of bones and related tissues such as cartilages
are investigated during their movements. For this reason, it is usually needed to have
an estimation of the joint center of rotation in advance. There are several methods
for estimating joint center of rotation, and different simulations may apply different
methods for obtaining the joint center. Therefore, it is very important to know how
the results of simulation may vary based on the methods used for estimating the joint
center of rotation. Arbabi et al. have evaluated this issue for hip joints by calculating
the virtual penetration depth during hip movement [ 13 , 16 ].
8.5.1 Hip Joint Center of Rotation
Many different methods of estimating hip joint center of rotation (HJC) have been
proposed that can be classified in predictive and functional approaches. The pre-
dictive (static) approach relies upon the location of anatomical landmarks [ 32 - 35 ].
The functional (dynamic) approach estimates the HJC from recorded [ 11 , 36 - 39 ]
or simulated [ 7 , 40 , 41 ] motion. Gilles depicted three predictive and two functional
approaches [ 40 ]. They are described here briefly.
8.5.1.1 Predictive Approaches
TheHJC is estimated as the center of the sphere that approximates the best the femoral
head or the acetabulum. The approximation is thus a least square fitting, which aims
to find the center (and radius) of the fitted sphere to the reconstructed data. These
two methods are denoted as femoral head sphere and acetabulum sphere methods.
The double sphere approach considers the joint as a perfect ball- and socket-joint in
which inter-articular distance is constant. It aims at finding the common center of
the femoral and acetabulum spheres [ 13 , 16 , 40 ].