Objective. The objective of this study is to illustrate the low accuracy of
two-dimensional (2-D) X-ray projection methods for the quantification of t
he three-dimensional (3-D) shoulder motions.
Background. The traditional method for the quantification of the gleno-hume
ral motion is by means of 2-D X-ray recording. The motion was characterized
by the scapulo-humeral rhythm: the ratio of the nett humeral elevation ove
r nett scapular rotation. The method was based on the quantification of the
planar projection of the spatial positions of X-ray dense structures of th
e scapula. The deformations introduced by the central projection method, a
feature of X-ray projection, cannot be compensated for by calibration: the
position of the scapula with respect to the camera setting is unknown, and
skeletal landmarks of the scapula cannot uniquely be identified. The transf
ormation from 3-D orientations to 2-D angles will, therefore, be inaccurate
.
Methods. A 2-D X-ray projection of the scapula during a typical arm abducti
on was simulated. The 3-D motion was obtained by means of palpation and sub
sequent digitization of skeletal landmarks of the scapula. The 3-D position
s of the recorded landmarks were projected on a plane by a simulation based
on the parameters of the X-ray equipment. The scapulo-humeral rhythm was c
alculated for the different scapular landmarks, and for the orientation of
the subject with respect to the projection axis. The results were compared
with previous published scapulo-humeral rhythms.
Results. The scapulo-humeral rhythm depends both on the choice of the skele
tal landmarks, used to quantify the scapular rotations, and on the orientat
ion of the subject in the X-ray setting. The full range of results obtained
from earlier published experiments could be obtained from a simulation bas
ed on a single 3-D arm abduction.
Conclusions. The 2-D scapulo-humeral rhythm, obtained from planar X-ray pro
jection, is an inaccurate parameter to define the scapular motions.