Date Published: February 11, 2019
Publisher: Public Library of Science
Author(s): Edin K. Suwarganda, Laura E. Diamond, David G. Lloyd, Thor F. Besier, Ju Zhang, Bryce A. Killen, Trevor N. Savage, David J. Saxby, Marco Livesu.
Accurate representation of subject-specific bone anatomy in lower-limb musculoskeletal models is important for human movement analyses and simulations. Mathematical methods can reconstruct geometric bone models using incomplete imaging of bone by morphing bone model templates, but the validity of these methods has not been fully explored. The purpose of this study was to determine the minimal imaging requirements for accurate reconstruction of geometric bone models. Complete geometric pelvis and femur models of 14 healthy adults were reconstructed from magnetic resonance imaging through segmentation. From each complete bone segmentation, three sets of incomplete segmentations (set 1 being the most incomplete) were created to test the effect of imaging incompleteness on reconstruction accuracy. Geometric bone models were reconstructed from complete sets, three incomplete sets, and two motion capture-based methods. Reconstructions from (in)complete sets were generated using statistical shape modelling, followed by host-mesh and local-mesh fitting through the Musculoskeletal Atlas Project Client. Reconstructions from motion capture-based methods used positional data from skin surface markers placed atop anatomic landmarks and estimated joint centre locations as target points for statistical shape modelling and linear scaling. Accuracy was evaluated with distance error (mm) and overlapping volume similarity (%) between complete bone segmentation and reconstructed bone models, and statistically compared using a repeated measure analysis of variance (p<0.05). Motion capture-based methods produced significantly higher distance error than reconstructions from (in)complete sets. Pelvis volume similarity reduced significantly with the level of incompleteness: complete set (92.70±1.92%), set 3 (85.41±1.99%), set 2 (81.22±3.03%), set 1 (62.30±6.17%), motion capture-based statistical shape modelling (41.18±9.54%), and motion capture-based linear scaling (26.80±7.19%). A similar trend was observed for femur volume similarity. Results indicate that imaging two relevant bone regions produces overlapping volume similarity >80% compared to complete segmented bone models, and improve analyses and simulation over current standard practice of linear scaling musculoskeletal models.
Musculoskeletal (MSK) lower-limb models are ubiquitous tools used in motion analysis and simulation. Individual variation in bone anatomy influences muscle attachments (i.e. origins and insertions), and thus muscle-tendon-unit paths and moment arms. Consequently, individual variation in bone anatomy influences estimates of muscle-tendon-unit forces, joint contact forces [1,2], and articular mechanisms  in these models. Furthermore, combined variation in bone anatomy and muscle-tendon-unit force can influence finite element analysis of bone , cartilage [5,6], and tendon . Therefore, generating accurate geometric bone models for subject-specific MSK models is important for human movement analysis and simulation.
This study was approved by Griffith University human research ethics committee (PES/36/10/HREC). All participants provided written informed consent prior to testing. Fourteen healthy adults (age = 28.79±4.82 years, height = 1.73±0.85 m, mass = 69.17±16.22 kg), who reported no history of lower-limb injury and were free from cardiovascular and neuromuscular conditions, participated.
Distance errors of geometric bone models reconstructed from (in)complete sets of bone segmentation through the MAP Client morphing process were significantly lower than MOCAP-based reconstructions of both pelvis and femur (Table 1). The distance error of reconstructed bone models from MOCAP-based SSM was also significantly lower than MOCAP-based linear scaling. Notably, femur models reconstructed from complete sets of segmentation had significantly higher distance error compared to reconstructions from incomplete sets 3.
The purpose of our study was to compare the accuracy of bone geometric models mathematically reconstructed from different levels of imaging incompleteness, and to identify the minimal imaging requirements for accurate representation of subject-specific bones. Our hypothesis, that geometric bone models reconstructed from more complete segmentations are more accurate than those reconstructed from less complete or MOCAP-based data, was partially supported. We found that pelvis and femur bone models reconstructed from different levels of imaging incompleteness had significantly different volume similarities compared to complete bone segmentations, but not significant different distance errors. Results suggest that imaging of only 2 bone regions, as represented by set 2 in this study, is sufficient (>80% volume similarity) to accurately reconstruct subject-specific pelvis and femur bone models.
Volume similarity of reconstructed geometric pelvis and femur models reduces as imaging becomes more incomplete. Using the MAP Client morphing process (i.e. SSM, followed by host-mesh, and then local-mesh fitting), medical imaging of only 2 relevant and truncated bone regions (set 2) is sufficient to accurately reconstruct both pelvis and femur bone models of healthy adults. In the absence of MRI, MOCAP-based SSM in the MAP Client is superior to MOCAP-based linear scaling of pelvis and femur bone models. These findings suggest that input of incomplete imaging and application of morphing techniques would improve analyses and simulation with adult subject-specific MSK models compared to standard practice of linear scaling generic MSK models.