A three-dimensional musculoskeletal model of the dog

Abstract

The domestic dog is interesting to investigate because of the wide range of body size, body mass, and physique in the many breeds. In the last several years, the number of clinical and biomechanical studies on dog locomotion has increased. However, the relationship between body structure and joint load during locomotion, as well as between joint load and degenerative diseases of the locomotor system (e.g. dysplasia), are not sufficiently understood. Collecting this data through in vivo measurements/records of joint forces and loads on deep/small muscles is complex, invasive, and sometimes unethical. The use of detailed musculoskeletal models may help fill the knowledge gap. We describe here the methods we used to create a detailed musculoskeletal model with 84 degrees of freedom and 134 muscles. Our model has three key-features: three-dimensionality, scalability, and modularity. We tested the validity of the model by identifying forelimb muscle synergies of a walking Beagle. We used inverse dynamics and static optimization to estimate muscle activations based on experimental data. We identified three muscle synergy groups by using hierarchical clustering. The activation patterns predicted from the model exhibit good agreement with experimental data for most of the forelimb muscles. We expect that our model will speed up the analysis of how body size, physique, agility, and disease influence neuronal control and joint loading in dog locomotion.

Figure 1

Abstraction of the forward and inverse simulation parameter chain for forelimb locomotion. Depending on the direction of the examination, the chain starts on the left or right side. The figure was created with the software package LibreOffice.

Figure 2

Forelimb calculated torques for the flexion/extension, abduction/adduction, and axial rotation in a walking Beagle based on consecutive strides of the same trial. Comparison between results obtained from OpenSim versus Newton–Euler method. The standard deviation (SD) is shown as shaded bands. Flexion/extension torques: positive values indicate net retractor torques and negative values indicate net protractor torques. A retractor torque flexes the shoulder joint, extends the elbow joint, and flexes the carpal joint. Abduction/adduction torques: negative values indicate abductor torque and positive values indicate adductor torque. Axial rotation torques: negative values indicate external rotation torques and positive values indicate internal rotation torques. The figures were created with the software package R.

Figure 3

Simulated forelimb muscle activation in a walking Beagle during one gait cycle, shown as a heatmap of logarithmic values (log2). The plot shows how individual muscles were activated based on consecutive strides of the same trial. The muscle groups (colours) were arranged according to hierarchical clustering (method—ward.d2) and minimal leaf sorting. The figures were created with the software packages OpenSim^, and R.

Figure 4

Hierarchical clustering (method—ward.d2) and minimal leaf sorting of simulated forelimb logarithmic muscle activation (log2) in the walking Beagle. The dendrogram shows how individual muscles were activated based on consecutive strides of the same trial. Groups represent a distance between activation patterns. The distance is displayed in the dendrogram as branch length. The longest lengths from the root were used as criteria to separate groups. The figure was created with the software package R.

Figure 5

High-resolution computed tomography (CT) data set of an anaesthetized Beagle (upper image—lateral view) and the reconstruction of the separated bones (lower image—lateral view) using the software package Amira.

Figure 6

Topology of the segments (boxes) and joints (arrows) of the whole Beagle (BE) model using the software package OpenSim^,.

Figure 7

Representation of the model assembly, from the bone model (yellow bones) to the muscle model (red paths) to the resulting simulation model, taking into account the transformations performed. The muscles (red) can be generated as paths closer to the real curves or just straight. Rotation around the red axis (x) represents protraction/retraction or flexion/extension, around the yellow axis (y) abduction/adduction, and around the green axis (z) axial rotation. The sub-figures were created using the software packages Amira and OpenSim^,.