Poster Presentation 50th International Society for the Study of the Lumbar Spine Annual Meeting 2024

Introduction:
Most spine models belong to either the musculoskeletal multibody (MB) or the finite element (FE) method. Coupling of MB and FE models is increasingly used to combine the advantages of both methods and to cover a more clinically relevant range of use cases. Active hybrid FE-MB models avoid the interface and convergence problems associated with model coupling but are still rarely used in spine research. Rather, they can provide the inherent ability to consider the full interaction of passive and active mechanisms for spinal stability. We therefore developed and validated a completely new model of the lumbosacral spine (LSS) in ArtiSynth¹ to identify the advantages and disadvantages of a forward dynamic active hybrid approach and to demonstrate new possibilities for its use.

Methods:
The model consists of the rigid vertebrae L1-S1 interconnected with hyperelastic fiber-reinforced FE intervertebral discs, ligaments, facet joints (Fig.1A), and force actuators representing the muscles. Thorax, pelvis, and auxiliary bodies are used to describe non-linear muscle paths and muscle attachments (Fig.1B). Kinematics of the abdominal plate is optimized using motion capture data and intra-abdominal pressure is calculated from the forces of the abdominal muscles compressing the abdominal cavity. Redundant muscle activations are predicted via a forward dynamics assisted data tracking to balance and move the LSS in different postures (Fig.1C). To increase the model validity and to utilize more data, we initially compared only the kinematic and structural mechanic responses of the passive LSS with in vitro data². For validation of the entire active hybrid model³, load cases derived from in vivo studies were simulated.

Results:
The developed LSS model (Fig.1) provides a robust and efficient estimation of biomechanical responses under in vivo similar loads, simultaneously calculating muscle activation patterns, internal mechanical loads, and vertebral movements without prescribing accurate spinal kinematics. Overall, all model responses, including range of motions, intervertebral rotations, intradiscal pressures, facet joint contact forces, instantaneous centers of rotation, functional spinal unit stiffnesses, intervertebral disc bulges, and intra-abdominal pressure, show high agreement with experimental values.

Discussion:
This work is a new approach to computational spine biomechanics that combines optimization-driven trunk musculature, FE intervertebral discs, ligaments, and facet joints in one model. The resulting capability to use the interplay of passive and neurally coordinated active mechanisms to simulate spinal stability represents an advance on the state-of-the-art. Our findings motivate a future application to develop patient-specific and pathological active hybrid spine models to study more complex load cases. This can be of therapeutic interest to identify conditions with great deflections but little stress on the intervertebral discs. More individualized therapies for muscle disorders through targeted strengthening or unloading are also conceivable. Replacing rigid vertebrae with FE bone structures is possible and can allow more comprehensive approaches to osseointegration and adjacent segment disease after surgical procedures. However, an important field of research remains the understanding and correlation to clinically relevant pain conditions.

Fig.1 – Passive components (A) of the developed active hybrid FE-MB LSS model (C,D).