Image-Based Proximity Mapping to Determine Joint Surface

Image-Based Proximity Mapping to Determine Joint Surface Interactions at the Elbow
1,2
Lalone, E A; 2McDonald, C P; 1,2Ferreira, L F; 1,2King, G J; +1,2 Johnson, J A
+1University of Western Ontario, London, ON,+2Hand and Upper Limb Centre, Bioengineering Laboratory, St. Joseph’s Healthcare, London, ON, Canada
[email protected]
INTRODUCTION:
The IBD algorithm was used to examine the effect of elbow flexion
Identifying articular contact patterns is important for both the
on ulnohumeral joint congruency. A view of the anterior surface of the
biomechanical investigation of joint mechanics as well as studying
humerus (with the ulna removed) is shown in Figure 2. At all angles of
osteoarthritis. Various in vitro methods have been employed to elucidate
flexion, the regions of closest joint proximity were concentrated on the
contact within diarthrodial joints including casting, staining and
medial side of the humerus. Throughout flexion, these regions of closest
interposed pressure films/sensors1-2 Although accurate, these methods
proximity migrated from the interior to the superior regions of the
are invasive requiring direct exposure of the joint.
humerus.
Recently, computed tomography (CT) and magnetic resonance
imaging (MRI) based approaches have been developed to non-invasively
Experimental Casting
Minimum Distance
3
quantify articular surface interactions as an estimate of joint contact .
Proximity Map
Despite the three-dimensional (3D) volumetric data acquired from these
Medial
Lateral
medical imaging modalities, these studies typically examine the joint in
a two-dimensional fashion4. Hence, the purpose herein was to develop a
proximity mapping technique to determine joint surface interactions,
using (3D) reconstructions of the joint structures. Validation employed
Coronoid
an experimental casting technique. The hypothesis was that proximity
maps of articulating surfaces would coincide with areas of joint contact
identified using the casting approach.
METHODS:
Algorithm Development:
Measuring articular cartilage cannot be done accurately using
A
B
computed tomography (CT). However, the location of underlying
Figure 1: Visual comparison of ulnohumeral joint contact using the (a)
subchondral bone can be defined and used to infer knowledge about the
experimental casting and (b) IBD algorithm.
articular cartilage. Therefore, joint surface interactions were modeled by
Effect of Elbow Flexion:
using volumetric images generated by CT. The proximity of the
subchondral bony surfaces of the joint was measured from 3D
reconstructions generated from CT slices using the Marching Cubes
algorithm available within the Visualization Toolkit. The Inter-bone
Superior
Distance (IBD) algorithm was developed by employing the vertices of
the polygons used to generate the surface, and these were used as
landmarks from which the algorithm searched for minimum distances
Lateral
Medial
between the two bony surfaces.
Experimental Techniques:
A denuded fresh-frozen ulnohumeral joint (61, F) was used for
validation. The ulnar and humeral shafts were positioned at 90° of
Interior
flexion and potted using dental cement. The specimens were then
Figure 2: Anterior humerus showing regions of articular joint contact
positioned in a custom designed CT compatible loading jig. A dental
throughout elbow flexion.
casting material was injected between the separated articular surfaces
DISCUSSION:
and (100N) of axial compression was applied (Reprosil Medium Body
The current study presents a novel approach for modeling joint
Vinyl Polysiloxine Impression Material, DENTPLY International Inc. –
surface interactions using proximity mapping calculated from imageYork, PA, USA). Subsequent to obtaining the joint cast, the elbow was
based measurements of subchondral bone distance. This technique was
imaged using a using a 64-slice CT scanner (GE LightSpeed VCT 64
validated using experimental casting to verify that it is possible to locate
Slice CT Scanner, New Berlin, WI) with the jig under the same loading
regions across the articulating surface that are most likely to be in
conditions as described above. 3D reconstructed surfaces of the humerus
contact. Using this approach, a contour map of proximity was obtained
and ulna were obtained using the volumetric image data from the CT.
to produce an inferred contact pattern across the articulating surfaces.
Proximity maps were generated using the IBD algorithm and then
Data was shown for simulated elbow flexion to illustrate this techniques'
compared with the casting approach.
capabilities as it determines regions of joint surface interactions using
Application of the IBD Algorithm to Examine Elbow Flexion:
the proximity maps throughout motion.
A single fresh-frozen upper extremity was used (48, M). A CT scan
Previous studies have investigated ulnohumeral contact report similar
of the frozen specimen was taken. After thawing, cables were sutured to
results as the current report. Goto et al5.determined proximity maps
the tendons of the brachialis; the biceps and triceps for loading. The
using magnetic resonance imaging to predict contact patterns. The
specimen was placed in a custom designed positioning device. CT data
contact pattern on the trochlear surface was also situated on the medial
was collected with the arm at 0˚, 30˚, 60˚ and 90˚ of elbow flexion while
facet of the trochlea, similar to the current study. Shiba et al.6 also
loading the biceps (44N) and brachialis and triceps (44N combined).
reported
that the contact pattern at the ulnohumeral joint migrates to the
A 3D reconstruction of the humerus and ulna was obtained at four
anterior-superior region of the humerus with increasing elbow flexion.
different angles of flexion. The IBD algorithm, measuring 3D surface to
Future work using this proximity mapping technique will examine joint
surface joint distances was applied and a proximity colormap displaying
surface interactions throughout a range of active elbow flexion using a
the relative distance between the two subchondral bone surfaces was
motion simulator previously developed in our laboratory. The clinical
then obtained.
implications of using imaging to determine joint contact interactions,
RESULTS:
especially throughout a range of motion, are significant as they should
Validation:
provide a useful clinical tool for patient evaluation when performed inThe cast of the ulnohumeral joint in the resulted in a five facet
vivo.
contact pattern. The pattern of the cast (Figure 1) closely resembled the
REFERENCES
proximity map generated using the inter-bone distance algorithm. Both
(1)Stormont T.J. et al. (1985) J Biomech.. 18(5), 329-36.
the casting and proximity maps revealed a large upper and lower medial
(2)Ronsky J.L. et al. (1995) J Biomech. August; 28(8), 977-83.
region of possible ‘contact’ which occurred on the periphery of the joint.
(3)Marai G.E. et al. (2004) IEEE Trans Biomed Eng. May; 51(5), 790-9.
Additionally, a large region of near joint ‘contact’ was found on the
(4)Gold G.E. et al. (2004) J Magn Res Imaging. Sept; 20(3), 526-30.
lateral side of the ulna located near the olecranon process which
(5)Goto A. et al. (2004) J Shoulder Elbow Surg. Jul-Aug; 13(4), 441-7.
extended down the lateral side of the ulna.
(6)Shiba R. et al. (1988) J Orthop Res. 6(6), 897-906.
Poster No. 1861 • 56th Annual Meeting of the Orthopaedic Research Society