Biomedical Engineering Reference
In-Depth Information
8.4 Joint Range Finder
8.4.1 Overview
Clinical hip examinations are usually based on rotating the hip in special orienta-
tions. For instance, the flexion-adduction-internal rotation test is used to aid in the
diagnosis of femoroacetabular impingement [
1
]. Such methods are usually based on
the patient's feedback during the examinations and therefore the diagnosis may not
be easy and accurate. After diagnosing the hip disease, the treatment may be based
on surgery. For example, options for treatment of femoroacetabular impingement
include trimming of the anterior aspect of the acetabular rim [
31
]. Since the opera-
tion can be highly invasive, it is essential that the surgeon has a good knowledge of
the joint's range of motion before the operation to know exactly the surgery strategy
and reduce the risk of miss-operation [
3
,
31
].
Not many studies have been performed on finding the joint's range of motion. The
current methods used for finding range of motion in human joints are usually based
on performing successive rotation increments as long as no collision occurs between
the 3D meshes of the fixed part and the mobile part along the selected anatomical
axis of rotation [
7
]. This can be highly time-consuming due to the fact that we cannot
know the number of rotation steps in advance. For increasing the simulation speed
sometimes the collision detection is restricted to certain area of the tissues. Imposing
such restrictions needs defining different areas of the tissues for the simulation pro-
gram [
7
], which is not easy to be done automatically. These traditional methods can
take a large amount of computing time. Arbabi et al. [
3
] proposed a method for find-
ing the maximum range of motion for human joints with rotating movements, based
on classifying the fixed part of the joint in a cylindrically segmented space (simi-
lar to the pre-processing step of cylindrical segmenting collision detection method,
explained in Sect.
8.2.2
), without using any collision detection algorithm. It proved
to be much faster than traditional ones, and needs to be performed only once per axis
of orientation [
3
,
31
].
8.4.2 Method
Similar to the pre-processing step of cylindrical segmenting collision detection
method, the algorithm starts by discretizing the space in the cylindrical orientation,
in the same orientation of rotation [
3
,
14
,
31
]. Then one table is prepared, with a
size that is function of the search interval and of the resolution. Each cell of the table
corresponds to one ring of the space and the indices of the fixed polygons occupying
that ring are stored in the corresponding table cell. For every vertex of the mobile
object, the corresponding ring-segment which contains the vertex is found. Finally,
the angular distance between the mobile vertex and the fixed polygons are calculated
in the corresponding table cell of the ring segment. By evaluating all ring segments,
