Femur, tibia and fibula bone templates to estimate subject-specific knee ligament attachment site locations

• Published in: Journal of biomechanics Vol. 49; no. 14; pp. 3523 - 3528
• Main Authors: Pillet, Hélène, Bergamini, Elena, Rochcongar, Goulven, Camomilla, Valentina, Thoreux, Patricia, Rouch, Philippe, Cappozzo, Aurelio, Skalli, Wafa
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 03.10.2016; Elsevier Limited; Elsevier
• Subjects: Accuracy; Aged; Algorithms; Attachment; Bone templates; Bones; Centroids; Collateral Ligaments - physiology; Estimates; Femur; Femur - physiology; Fibula - physiology; Generic model; Humans; Knee; Knee Joint - physiology; Knee ligaments; Knees; Life Sciences; Ligament footprint; Ligaments; Medical imaging; Middle Aged; Models, Biological; Morphology; NMR; Nuclear magnetic resonance; Physical Medicine and Rehabilitation; Scaling; Studies; Surgery; Tibia; Tibia - physiology; Generic model; Scaling; Ligament footprint; Knee ligaments; Bone templates
• ISSN: 0021-9290; 1873-2380; 1873-2380
• DOI: 10.1016/j.jbiomech.2016.09.027

In-vivo estimates of the positions of knee ligament attachment sites are crucial for subject-specific knee modelling. The present study provides template digital models of femur, tibia and fibula that embed the positions of centroids of the origins and insertions of cruciate and collateral ligaments, along with information on their dispersion related to inter-individual variability. By using a shape transformation procedure of choice, these templates can be made to match anatomical information measured on a subject under analysis. Generic bone digital models of the femur, tibia and fibula were first chosen as bone templates. Ligament attachment areas were accurately identified through dissection on the bones of 11 knee specimens, and marked using radio opaque paint. Digital models of these bones embedding the positions of the centroids of the identified ligament attachment areas were thereafter obtained using medical imaging techniques. These centroids were mapped onto the relevant bone template, thus obtaining a cloud of 11 points for each attachment site, and descriptive statistics of the position of these points were thereafter determined. Dispersion of these positions, essentially due to inter-individual variability, was below 6mm for all attachment areas. The accuracy with which subject-specific ligament attachment site positions may be estimated using the bone template models provided in this paper was also assessed using the above-mentioned 11 specimens data set, and a leave-one-out cross validation approach. Average accuracy was found to be 3.3±1.5mm and 5.8±2.9mm for femoral and tibial/fibular attachment sites, respectively.