A thoracolumbar multibody model capable of simulating scoliosis deformities

INTRODUCTION
Adolescent Idiopathic Scoliosis is identified as a multifactorial disease, which the aetiopathogenesis behind it has remained controversial. We hypothesize that musculoskeletal modeling can offer an insight into the matter. This work presents an improved and kinematically determinate thoracolumbar spine model capable of reconstructing large spine deformations.

METHOD
The model was created using the AnyBody Modeling System (AMS, v. 7.2). The previously presented lumbar spine model (De Zee et al. 2007) together with a thoracolumbar spine model with articulated ribcage (Ignasiak et al. 2016) form the basis for the development of the new multibody spine model. Scoliosis deformities and severity can be described by clinically accepted measures and anatomical degrees-of-freedom. These scoliosis variables such as calculated after placement of anatomic landmarks at osseous elements on medical imaging data. Several modifications of joint definitions compared to previous models were performed. These modifications led to a kinematically determinate model that is drivable by clinically accepted measures. The proposed thoracic model, including T10 to T1 vertebrae, non-floating ribs, and sternum, has 17 DOF after implementation of the kinematic constraints.

RESULT AND DISCUSSION
The model appears to reproduce scoliosis variables according to typical deformations for scoliosis, thus verifying qualitatively that the model constraints are correct. Forthcoming work will attempt quantitative verification of the deformation patterns. Correct kinematic constraints are a condition for the subsequent use of the model to investigate the kinetics of scoliosis aetiology.

CONCLUSION
The proposed thoracolumbar model is able to simulate scoliosis deformities and enhance state-of-the-art on scoliosis aetiology as well as biomechanics of the torso.