3D video-based deformation measurement of the pelvis bone under dynamic cyclic loading
© Göpfert et al; licensee BioMed Central Ltd. 2011
Received: 19 January 2011
Accepted: 17 July 2011
Published: 17 July 2011
Dynamic three-dimensional (3D) deformation of the pelvic bones is a crucial factor in the successful design and longevity of complex orthopaedic oncological implants. The current solutions are often not very promising for the patient; thus it would be interesting to measure the dynamic 3D-deformation of the whole pelvic bone in order to get a more realistic dataset for a better implant design. Therefore we hypothesis if it would be possible to combine a material testing machine with a 3D video motion capturing system, used in clinical gait analysis, to measure the sub millimetre deformation of a whole pelvis specimen.
A pelvis specimen was placed in a standing position on a material testing machine. Passive reflective markers, traceable by the 3D video motion capturing system, were fixed to the bony surface of the pelvis specimen. While applying a dynamic sinusoidal load the 3D-movement of the markers was recorded by the cameras and afterwards the 3D-deformation of the pelvis specimen was computed. The accuracy of the 3D-movement of the markers was verified with 3D-displacement curve with a step function using a manual driven 3D micro-motion-stage.
The resulting accuracy of the measurement system depended on the number of cameras tracking a marker. The noise level for a marker seen by two cameras was during the stationary phase of the calibration procedure ± 0.036 mm, and ± 0.022 mm if tracked by 6 cameras. The detectable 3D-movement performed by the 3D-micro-motion-stage was smaller than the noise level of the 3D-video motion capturing system. Therefore the limiting factor of the setup was the noise level, which resulted in a measurement accuracy for the dynamic test setup of ± 0.036 mm.
This 3D test setup opens new possibilities in dynamic testing of wide range materials, like anatomical specimens, biomaterials, and its combinations. The resulting 3D-deformation dataset can be used for a better estimation of material characteristics of the underlying structures. This is an important factor in a reliable biomechanical modelling and simulation as well as in a successful design of complex implants.
Video motion capturing (MoCap) systems are widely used in animation industries and also in biomechanical applications with the main focus of macroscopic gait analysis . Due to systems improvements over the last years, they are now capable of resolutions of 4704 × 3456 pixels (T160, Vicon, Oxford, UK) and sampling rates up to 10000 frames per second (Raptor 4, Motion Analysis Corporation, Santa Rosa, CA, USA). This increased technology opens new application areas where high camera resolution is needed such as in measuring small deformations of biological tissues, dynamic three-dimensional (3D) deformation of bone-implant systems, or to determine movement after fracture fixation [2–6]. Further, micro motion between bone and implant is an important factor in the longevity of a stable bone-implant interface [7–9]. In particular, the characteristics of complex implant systems with bio-absorbable materials or bioactive surface coatings are not completely known . Variation in daily activities may alter the loading conditions in the bone-implant system because of changing material properties on one hand and the influences of operations on the other. However, biomechanical loading tests can never cover the whole variety of different loading conditions during daily activity or consider all the changes over time in the biological or implant structure . Although computation-based modelling tries to include in its calculation process changes of the bone-implant system; its original data must be based on real validated measurements otherwise it can lead to false conclusions [12, 13]. Nevertheless local deformation can be measured highly precise with strain gauges [14, 15], or linear variable differential transducers . To measure the 3D-deformation of a whole specimen other methods give a better spatial resolution. Therefore the hypothesis was that a three-dimensional video MoCap system has the potential to measure the sub millimetre 3D-deformation of a whole pelvis specimen.
This study describes the application of dynamic 3D-deformation measurement using a 3D video MoCap system to gain data for the development of complex orthopaedic oncological implants around the hip joint of the pelvis including the determination of its accuracy with a 3D micrometre stage. The measurement of 3D-deformation was done by recording the 3D-movement of passive reflective markers glued onto bone surface while applying a dynamic, cyclic, non-constrained, uni-axial load to the pelvic bone. The results will help achieve more stable implant fixation to the bone, which improves the initial conditions for successful osseointegration and therefore support the durability of the implant in a complex environment after a reconstructive surgery for bone tumour.
Determination of the accuracy of the 3D motion-capture system
Before performing the accuracy determination procedure, the cameras were calibrated as described in the Vicon Handbook using the 120 mm 3-marker calibration wand and the Ergo-Calibration Frame equipped with 9 mm and 9.5 mm markers respectively. The sampling rate of the cameras was set at 60 Hz and the hydraulic pump of the Servo-hydraulic testing machine was running in order to get conditions identical to the 3D-deformation measurements with the pelvic specimen. The recorded 3D video data of the reflecting markers were tracked using the tracking software Vicon Nexus (Vicon, Oxford, UK) without using any filter function and then exported in to Microsoft Excel (Microsoft Corporation, Redmond, WA, USA). The 3D accuracy of the spatial resolution of the each marker was determined during the three middle seconds of the stationary phase after each 3D-displacement step. A simultaneous calculation of the three static markers was done.
Test procedure with the pelvic specimen
Data processing of the 3D marker data of pelvic specimen
The recorded 3D video data of the reflecting markers were reconstructed and tracked using the Vicon Nexus software without using any filter function. Each tracked marker was seen simultaneously by a minimum of 3 cameras. The dataset of each loading condition consists of the global 3D coordinates of all the markers and the applied load. It was exported into a CSV file for further use like in graphical animation software or displacement computation.
Where Δd(t) is the temporal distance between the different spatial coordinates (xi(t); yi(t); zi(t)) of two markers over the loading cycles.
Accuracy of the 3D motion-capture system setup
Accuracy of the measurement setup depending on the number of cameras tracking a marker
Marker visible by n cameras
Noise level [+/- mm] at 0.01 mm step procedure
Noise level [+/- mm] at 0.1 mm step procedure
Examples of 3D-displacement and 3D-deformation between two markers
To illustrate the movement of the specimen and deformation of the pelvis, two pairs of markers were analysed. One pair was chosen at the iliac crest and another between the acetabulum and femur. The loading conditions were the same; a sinusoidal loading of 100 cycles at 1 Hz with a loading amplitude between 100 N and 900 N (1 × body weight). The 10th and 90th cycles were analysed.
The chosen system, combining 3D-video MoCap with servo-hydraulic material testing, allows dynamic 3D-displacement measurement of the reflecting markers placed on the surface of a specimen in the sub millimetre range. The accuracy of the measurable marker displacement with a 3D video MoCap system depends on the number of cameras tracking a single marker simultaneously, on the size and shape of the markers, and the quality of the camera (CCD-Chip, objective, camera fixation). The accuracy of ± 0.036 mm reached with the setup used in this study is in the range published by Windolf et al. , or by Lujan . The setup used by Liu et al.  has a higher accuracy but also a smaller measurement volume. The accuracy could be increased by more cameras tracking simultaneous each markers, cameras with a higher resolution and a lower noise level or with bigger markers .
The advantage of bigger markers is that there is better reflection of the camera light source at the reflecting marker surface towards the detecting CCD-chip due to smaller curvature of the surface . On the other hand, bigger markers lead to bigger inter-marker distance, or increase the risk of covering some markers due to the bigger volume. A bigger marker distance or less visible markers would lead to a reduction in resolution of the deformation measurement. However, smaller markers could increase the resolution of the deformation due to the smaller inter-marker distance but would be more difficult to handle and place on the surface of the specimen. Additionally a too small the inter-marker distance increases also the risk that two markers merge into one reflecting spot. In this case, a marker would not be properly tracked and its 3D data would be lost. Therefore the dome-shaped markers with a diameter of 6.5 mm, and, based on experience, an inter-marker distance of at least two times the marker diameter are a good compromise between high accuracy and easy handling needed for gluing onto the bony surface.
It has to be taken in account that the marker displacement is usually a combination of the global movement of the specimen in space and deformation occurring through the loading process. However, the movement path of the markers during a loading cycle comprises all the information about the whole specimen and the local loading conditions. Nevertheless, it is possible to draw some conclusions about the bony structure of the specimen or the movement of a joint between different markers, without knowing the exact underline bony structure. Although the quality of the information behind the surface displacement data will be improved by additional non-destructive technologies like computed tomography (CT) scans, further improvements might be possible by local micro-CT scans .
A big benefit of the used plastic reflecting marker is that they don't induce any artifacts on the CT-images. Therefore it is relative easy to combine the 3D-deformation data with the CT-based bone-density and structure information. The spatial resolution (voxel size) of the current CT-scanner is in the range of 1 mm. That means the CT-scanner has about 20-times smaller resolution than the MoCap-system. However, this difference in resolution is not the limiting factor by combining the MoCap-system and the CT-scanner. The limiting factor to acquire a high precision resolution 3D bone deformation in connection with the CT-data of the underlining bony structure is actually the inter-marker distance of the reflecting markers as described above.
The combination of 3D video MoCap, and material testing opens new possibilities in dynamic testing. Combined with CT-data of the underlining bony structure, it becomes highly valuable framework for finite element modelling of complex implants . It may also improve the development process of new implant technologies through better biomechanical compatibility with the patient specific musculoskeletal anatomy.
Acknowledgements and Funding
This project was support by a grant from the ENDO-Stiftung, Hamburg, Germany. We thank Cora Huber, Corina Nüesch, Sarah Schelldorfer, and Dieter Wirz for their help during the measurements.
- Blake R, Ferguson H: The motion analysis system for dynamic gait analysis. Clinics in podiatric medicine and surgery 1993, 10–3: 501.Google Scholar
- Windolf M, Klos K, Wähnert D, Van Der Pol B, Radtke R, Schwieger K, Jakob R: Biomechanical investigation of an alternative concept to angular stable plating using conventional fixation hardware. BMC Musculoskeletal Disorders 2010, 11–1: 95.View ArticleGoogle Scholar
- Liu H, Holt C, Evans S: Accuracy and repeatability of an optical motion analysis system for measuring small deformations of biological tissues. J Biomech 2007, 40–1: 210–4.View ArticleGoogle Scholar
- Hirokawa S, Yamamoto K, Kawada T: Circumferential measurement and analysis of strain distribution in the human ACL using a photoelastic coating method. Journal of biomechanics 2001, 34–9: 1135.View ArticleGoogle Scholar
- Green T, Allvey J, Adams M: Spondylolysis: bending of the inferior articular processes of lumbar vertebrae during simulated spinal movements. Spine 1994, 19–23: 2683.View ArticleGoogle Scholar
- Häggman Henrikson B, Eriksson P, Nordh E, Zafar H: Evaluation of skin versus teeth attached markers in wireless optoelectronic recordings of chewing movements in man. Journal of oral rehabilitation 1998, 25–7: 527–34.View ArticleGoogle Scholar
- Kienapfel H, Sprey C, Wilke A, Griss P: Implant fixation by bone ingrowth. The Journal of arthroplasty 1999, 14–3: 355–68.View ArticleGoogle Scholar
- Tarala M, Janssen D, Telka A, Waanders D, Verdonschot N: Experimental versus computational analysis of micromotions at the implant-bone interface. Proceedings of the Institution of Mechanical Engineers, Part H: Journal of Engineering in Medicine 2010, 224: 1–8.Google Scholar
- Widmer K, Wu J, Zurfluh B, Gopfert B, Morscher E: Three-dimensional secondary stability of cemented and non-cemented acetabular implants ex-vivo under dynamic load. Journal of biomechanics 1998, 31–1001: 166.View ArticleGoogle Scholar
- Roach P, Eglin D, Rohde K, Perry C: Modern biomaterials: a review--bulk properties and implications of surface modifications. Journal of Materials Science: Materials in Medicine 2007, 18–7: 1263–77.Google Scholar
- Currey J: Bone strength: What are we trying to measure? Calcified Tissue International 2001, 68–4: 205–10.View ArticleGoogle Scholar
- Henninger H, Reese S, Anderson A, Weiss J: Validation of computational models in biomechanics. Proceedings of the Institution of Mechanical Engineers, Part H: Journal of Engineering in Medicine 2010, 224–7: 801–12.View ArticleGoogle Scholar
- Cristofolini L, Schileo E, Juszczyk M, Taddei F, Martelli S, Viceconti M: Mechanical testing of bones: the positive synergy of finite-element models and in vitro experiments. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences 2010, 368–1920: 2725.View ArticleGoogle Scholar
- Gray H, Zavatsky A, Taddei F, Cristofolini L, Gill H: Experimental validation of a finite element model of a composite tibia. Proceedings of the Institution of Mechanical Engineers, Part H: Journal of Engineering in Medicine 2007, 221–3: 315–24.View ArticleGoogle Scholar
- Anderson A, Peters C, Tuttle B, Weiss J: Subject-specific finite element model of the pelvis: development, validation and sensitivity studies. Journal of Biomechanical Engineering 2005, 127: 364. 10.1115/1.1894148View ArticleGoogle Scholar
- Abdul-Kadir M, Hansen U, Klabunde R, Lucas D, Amis A: Finite element modelling of primary hip stem stability: The effect of interference fit. Journal of biomechanics 2008, 41–3: 587–94.View ArticleGoogle Scholar
- Widmer KH, Zurfluh B, Morscher EW: Contact surface and pressure load at implant-bone interface in press-fit cups compared to natural hip joints. Orthopade 1997, 26–2: 181–9.Google Scholar
- Windolf M, Gotzen N, Morlock M: Systematic accuracy and precision analysis of video motion capturing systems--exemplified on the Vicon-460 system. J Biomech 2008, 41–12: 2776–80.View ArticleGoogle Scholar
- Lujan TJ, Lake SP, Plaizier TA, Ellis BJ, Weiss JA: Simultaneous measurement of three-dimensional joint kinematics and ligament strains with optical methods. J Biomech Eng 2005, 127–1: 193–97.Google Scholar
- Rho J, Hobatho M, Ashman R: Relations of mechanical properties to density and CT numbers in human bone. Medical Engineering & Physics 1995, 17–5: 347–55.View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.