Research into Spinal Deformities 8 T. Kotwicki and T.B. Grivas (Eds.

) IOS Press, 2012 2012 The authors and IOS Press. All rights reserved. doi:10.3233/978-1-61499-067-3-227

227

An improved kinematic model of the spine for three-dimensional motion analysis in the Vicon system
Mirosawa M. DUGOSZa,1, Daria PANEK a, Pawe MACIEJASZ b, Wiesaw CHWAA c, Witold ALDA a a AGH University of Science and Technology, Krakw b DEMAR - LIRMM, University of Montpellier 2, INRIA, CNRS, Montpellier, France c University School of Physical Education in Krakow
Abstract. The mechanism of creation and pathomechanics of lateral spinal deformation is still not fully explained. Modern medical imaging techniques give scientists possibility to understand some aspects, but vast majority of those techniques is based on static trials. A motion capture system belongs to techniques which enable visualization of a spine during dynamic trials; however, due to lack of appropriate computational model, it is unsuitable for scoliosis imaging. A few years ago our group has proposed a kinematic model of the spine to be used with Vicon Motion Capture System, which was based on Bzier curves. That model allowed for much more precise investigation of spinal kinematics during dynamic trials as compared with other computational models. However, it did not allowed to restrict only selected movements for particular segments of the spine (e.g. axial rotation for lumbar spine). The aim of the current work is to improve the proposed model in order to be able to restrict selected movements according to the knowledge concerning spinal anatomy and spinal range of motion. The new kinematic model of the spine was written in BodyBuilder for Biomechanics Language. For the purpose of visualization also an accurate graphical representation of each vertebra (polygon mesh) was computed and adapted to be compatible with the kinematic model. Using a new version of the model it is possible to perform precise analysis of movement of all vertebrae during such dynamic activities as e.g. gait and forward or lateral bending, as well as to present the results not only on the charts, but also as a 3D animation of movements of a realistically looking spine. The paper describes the new kinematic model and the process of creating graphical representation of the vertebrae. Also sample results obtained using that model are presented. Keywords. Vicon Motion Capture System, spine modeling

Introduction The mechanism of formation and pathomechanics of lateral spinal deformation is still not fully explained. Modern medical imaging techniques give scientists possibility to understand some aspects of it, but a vast majority of the techniques is based on static systems.
1

Corresponding Author, e-mail: mmd@agh.edu.pl

228

M.M. Dugosz et al. / An Improved Kinematic Model of the Spine

A human spine, due to the functions it performs, as well as due to frequent pathologies and injuries, for centuries has remained the center of attention in many fields of science and technology. It is the axis of the human skeleton, which carries the weight of the upper half of the body and protects the spinal cord. The etiology of various diseases and injuries of the spine, which can be harmful to human health, has resulted in the need for interdisciplinary research based on different measurement techniques used in biomedicine. Experiments, performed currently in research centers all over the world, allow precise observation of the properties and biomechanics of the human spine, and therefore are the main source of knowledge about it [1, 2, 3]. So far there is no precise mathematical model that could accurately represent the behavior of such a complex system, as the spine. We still do not know several mechanisms that cause abnormalities of the human body posture, making the selection of an appropriate rehabilitation and treatment very difficult. To solve that problem, in addition to static tests, it is necessary to analyze the dynamics of the musculoskeletal system. Widely used imaging techniques allow examining human body mainly statically or in a very limited range of motion, which often is not enough to take steps for further treatment. In addition, most of these methods use ionizing radiation which is harmful to the health of patients and therefore should be used rarely [3]. In the presented approach we use the Vicon Motion Capture System and a new model developed by our team to analyze and visualize human spine kinematics.

1. Material and Methods A few years ago our group has proposed a kinematic model of the spine based on Bzier curves, to be used with the Vicon Motion Capture System [4,5]. That model allowed much more precise investigation of spinal kinematics during dynamic trials as compared with other computational models. However, it did not allow to specify various ranges of motion in various planes for the same spinal segment. Therefore, using that model it was difficult to allow, e.g. significant flexion, and in the same time to restrict axial rotation in the lumbar region. The aim of our current work is to improve the proposed model in order to be able to examine selected movements according to the knowledge concerning spinal anatomy and spinal range of motion. The new kinematic model of the spine was written in the Vicons BodyBuilder for Biomechanics language. For the purpose of visualization an accurate graphical representation of each vertebra (represented by polygon mesh) was also computed and adapted to be compatible with the kinematic model [6]. To build this model, authors simultaneously used two techniques: 3D scanning of the vertebrae and computer modeling using 3D graphics software. In the BodyBuilder language, a human body is treated as a set of rigid bodies (segments) connected by hinges (joints). Each point is described in the global and local (segment's) coordinate system. The new model that we have developed is using information about the range of motion of each vertebra (in fact of each connection of two vertebrae) in the three planes separately. This information should be provided before the computation is started. In our model we are using the ranges of motion specified in [2]. In our model the exact position and orientation of each vertebra is calculated according to the following routine:

M.M. Dugosz et al. / An Improved Kinematic Model of the Spine

229

1.

Determine the location and the orientation of the segments C0 (occipital bone), T1 (T1 vertebra) and S1 (sacrum), based on the positions of the anatomical landmarks Calculate the rotation (orientation) of the C0 segment in the local coordinate system of the T1 segment, and the rotation of the T1 segment in the local coordinate system of the S1 segment Divide the rotation of the C0 segment determined in the step 2 between the vertebra C1 to T1, proportionally to their ranges of motion (provided as input parameters), separately in each plane. Perform similar operation for vertebrae T1 to S1 Compute the relative rotation of each pair of vertebrae, by summing their initial relative rotations in the axial plane (provided as input parameters) and the rotation determined in the step 3 Calculate distances between each two successive vertebrae based on the knowledge of relative heights of each vertebra (provided as input parameters), their current rotation (determined in step 4) and the current positions of C0, T1 and S1 segments Determine the current position and orientation of each vertebra in the global coordinate system.

2.

3.

4.

5.

6.

In the last step of the algorithm, the whole spine is built from the blocks, by calculating the current position and orientation of each vertebra one by one. First, the position and the orientation of the L5 vertebra is determined based on the position and the orientation of the S1 vertebra (step 1 of the algorithm) as well as on the knowledge of the orientation (step 4) and the position (step 5) of the L5 vertebra in the local coordinate system of the S1 segment. Then, the same operation is performed for the L4 segment, but this time in the relation to the L5 vertebra. This operation is repeated until the positions and the orientations of all the vertebrae in the global coordinate system are determined. The above sequence of operations is performed for each time frame.

2. Results So far the new model was verified on two subjects. In both cases the same sets of exercises was performed, i.e. forward, backward and lateral bending, as well as torso twist in the erect position and associated with forward bending. During this trials, additionally to C7 and T10 markers (used by both models for the computation), T3, T6, L1 and L4 markers were placed on the skin above the appropriate vertebra (see fig. 1) in order to verify the performance of the both models. The markers were mainly used to evaluate the curvature of the spine in the axial plane. During movement spine moves under the skin, therefore the positions of the markers on the skin do not always correspond to the positions of particular vertebrae under it. Thus the markers placed on the skin can not be used to verify, if the curvature of the spine in the frontal plane is correctly computed.

230

M.M. Dugosz et al. / An Improved Kinematic Model of the Spine

Figure 1. Visualization of the spine during three-dimensional motion computed using the proposed models. Left: spinal column reconstructed using the new version of the model. Right: the difference in the spinal curvatures computed using the previous (green) and the new (orange) version of the model during backward bending. In both figures the actual positions of the markers C7, T3, T6, T10, L1 and L4 attached to the skin of the patient are marked with white dots.

The visual comparison of the positions of the markers over spine with the positions of particular vertebrae computed using both versions of the model has revealed that in some cases (e.g. backward bending see figure 1 right) the new version of the model allows for better determination of spinal curvature. In some other cases (e.g. lateral bending) the results obtained by both models are very similar. Finally, in some cases (e.g. torso twists) the differences between spinal curvatures computed using both versions of the models are visible, however, it is not clear which results are correct. It is due to lack of reliable method allowing for determination of vertebrae positions during movement. The results provided by the both models depend not only on the trajectories of the markers registered by the motion capture system, but also on the values of some input parameters. Especially, the new version of the model uses a high number of such parameters, which are: relative heights of each vertebra (used also by the previous version of the model), ranges of motion of each vertebra and initial rotation of each vertebra in the axial plane. Thanks to those parameters the new model allow for higher level of adaption and therefore should be able to provide correct results irrespectively of the kind of performed motion. However, further investigation is necessary to determine optimal values of those parameters.

M.M. Dugosz et al. / An Improved Kinematic Model of the Spine

231

3. Conclusions The new kinematic and geometric model we have developed allows for accurate threedimensional analysis of the spinal cord movements. In particular it allows to precisely calculate the position and the orientation of each vertebra during each moment of the movement, therefore also the changes of spinal curvature may be visualized with a very good temporal resolution. It may also allow to track very precisely the angles between each pair of vertebrae during movement, thus being very useful tool for practitioners of spinal deformities rehabilitation. The model meets the design goals, while taking into account the ranges of movement specific to the individual vertebrae, maintaining at the same time the shape of the physiological curvature of the spine. Although the model allows for high level of adaptation, to make use of it, it is necessary to find a reliable method for validation of the obtained results. Also further investigations should be performed in order to determine the correct values of the input parameters. After successfully passed validation, the presented model could be used for diagnosing and monitoring the progress of treatment of various diseases and pathologies of the spine. Our team aims particularly at the use of the proposed model for the evaluation of scoliosis.

References
[1] M. Gzik, Identyfikacja si w strukturach anatomicznych krgosupa szyjnego czowieka , Gliwice, Politechnika lska, 2008 (in Polish). [2] A.A. White, M. Panjabi, Clinical Biomechanics of the Spine , wyd. 2. Toronto, Lippincott Williams & Wilkins 1990. [3] R. Tadeusiewicz, P. Augustyniak, Podstawy Inynierii Biomedycznej, tom I, Krakw, Wydawnictwo AGH, 2009. [4] P. Maciejasz, W. Chwaa, A method of evaluating the kinematics of a spine deformed by scoliosis using a motion capture system, Journal of Biomechanics, 39 (2006), 172. [5] P. Maciejasz, W. Chwaa, M. Biaek, T. Kotwicki, A method for investigation of spinal kinematics in children with idiopatic scoliosis, 5th International Conference on Conservative Management of Spinal Deformities, Athens, Greece. 35 April 2008, Scoliosis 4, Suppl. 1 (2009), O21; doi:10.1186/17487161-4-S1-O21 [6] M. Dugosz, W. Chwaa, P. Maciejasz, W. Alda, Realistic Model of Spine Geometry in the Human Skeleton in the Vicon System, Bio-Algorithms and Medical Systems 8(1), 2012.

Copyright of Studies in Health Technology & Informatics is the property of IOS Press and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.