173. 3
D imaging of the spine using EOS system, is it reliable? A
comparative study using CT imaging.
Zaid Al-Aubaidi , David Lebel , Rienhard Zeller , Kamaldine Oudjhane,
Benjamin Alman
Orthopaedic Odense Universitet hospital ; Orthopaedic , Sick Kids hospital,
Toronto, Canada; Orthopaedic, Sick Kids hospital, Toronto, Canada;
Orthopaedic, Sick Kids hospital, Toronto, Canada; Orthopaedic , Sick Kids
hospital, Toronto, Canada
Background:
EOS was developed from a Nobel Prize- winning technology by a team
of engineers; orthopaedic surgeons and radiologists. 3D views can be obtained from
EOS through special algorithms involving manual outlining of the spinal contours by
doing manual reconstructions. This process is based on the projection of elements out
of a 3D library onto the 2D spinal contours generated by orthogonal X-rays.
Purpose / Aim of Study:
The aim of this study is to evaluate the precision of the 3D
geometry.
Materials and Methods:
All patients presented with scoliosis, which had undergone
EOS spinal X-rays and CT scan for a better visualization of their spinal deformity. To
decrease the risk of different measurements due to progression of the deformity, the
time limit between performing the CT and EOS was set to be not more than 6 months.
The rotation was measured using Apical Vertebral Orientation (AVO) measured by the
EOS, and the same method was used to measure the orientation of the apical vertebra
by the CT. Due to the fact that CT is supine and EOS is standing, other measurements
as the Cobb angle and the Apical Vertebral Rotation (AVR) according to Nash and
Moe, were used. These were considered as the control variables measured using the
EOS and the CT scout view, as to control the changes between the standing in EOS vs.
the supine position in CT scanning. To evaluate the intraobserver and interobserver
error regarding the precision of AVO measurements, these were repeated by the first
author and the second author. For statistical analysis the Graph Pad soft ware was used
and the unpaired t test was used.
Findings / Results:
7
patients, 3 males and 4 females, age between 10-18 years
(15
y5m). 3 patients were diagnosed with kyphoscoliosis, one with neurofibromatosis
type I, one with Marfan syndrome, one with Myelomeningocelle and one with
lymphoma. The average Cobb angle measured using the EOS was 32,25 (9-78), while
using the CT scout view the Cobb angle was 26.38 (P= 0,56). There were no
statistically significant different regarding the inter- and intraobserver error. The
average of these measurements was 8,3 vs. the measurement done by the CT 6,6 (P=
0,65).
Only one of seven differed in the two AVR assessment according to Nash and
Moe (p=0.29Fisher Exact).
Conclusions:
The precision of EOS based measurement of vertebral rotation has never
been tested in clinical practice. To the best of our knowledge this is the first study that
looks at the precision of the EOS system compared to measurements obtained by CT.
Our study did not find any statistically significant difference between EOS and CT
vertebral rotation measurements. Although has limitations, this study does conclude
that the measurement obtained by the EOS is comparable to that obtained by CT.
