Replicating spine loading during functional and daily activities: An in vivo, in silico, in vitro research pipeline.

Lifestyle heavily influences intervertebral disc (IVD) loads, but measuring in vivo loads requires invasive methods, and the ability to apply these loads in vitro is limited. In vivo load data from instrumented vertebral body replacements is limited to patients that have had spinal fusion surgery, potentially resulting in different kinematics and loading patterns compared to a healthy population. Therefore, this study aimed to develop a pipeline for the non-invasive estimation of in vivo IVD loading, and the application of these loads in vitro. A full-body Opensim model was developed by adapting and combining two existing models. Kinetic data from healthy participants performing activities of daily living were used as inputs for simulations using static optimisation. After evaluating simulation results using in vivo data, the estimated six-axis physiological loads were applied to bovine tail specimens. The pipeline was then used to compare the kinematics resulting from the physiological load profiles (flexion, lateral bending, axial rotation) with a simplified pure moment protocol commonly used for in vitro studies. Comparing kinematics revealed that the in vitro physiological load protocol followed the same trends as the in silico and in vivo data. Furthermore, the physiological loads resulted in substantially different kinematics when compared to pure moment testing, particularly in flexion. Therefore, the use of the presented pipeline to estimate the complex loads of daily activities in different populations, and the application of those loads in vitro provides a novel capability to deepen our knowledge of spine biomechanics, IVD mechanobiology, and improve pre-clinical test methods.

The physiological, in-vitro simulation of daily activities in the intervertebral disc using a load Informed kinematic evaluation (LIKE) protocol.

Current spinal testing protocols generally adopt pure moments combined with axial compression. However, daily activities involve multi-axis loads, and multi-axis loading has been shown to impact intervertebral disc (IVD) cell viability. Therefore, integrating in-vivo load data with spine simulators is critical to understand how loading affects the IVD, but doing so is challenging due to load coupling and variable load rates. This study addresses these challenges through the Load Informed Kinematic Evaluation (LIKE) protocol, which was evaluated using the root mean squared error (RMSE) between desired and actual loads in each axis. Stage 1 involves obtaining the kinematics from six-axis load control tests replicating 20 Orthoload activities at a reduced test speed. Stage 2 applies these kinematics in five axes, with axial compression applied in load control, at the reduced speed and at the physiological test rate. Stage 3 enables long-term tests through six-axis kinematic control combined with diurnal height correction to account for the natural height fluctuations of the IVD. Stage 1 yielded RMSEs within twice the load cell noise floor. Low RMSEs were maintained during stage 2 at reduced speed (Tx:0.80 ± 0.30 N; Ty:0.77 ± 0.29 N; Tz:1.79 ± 0.50 N; Rx:0.02 ± 0.01Nm; Ry:0.02 ± 0.01Nm; and Rz:0.02 ± 0.01Nm) and at the physiological test rate (Tx:3.45 ± 1.81 N; Ty:3.82 ± 1.99 N; Tz:11.32 ± 8.69 N; Rx:0.13 ± 0.07Nm; Ry:0.16 ± 0.11Nm; and Rz:0.07 ± 0.04Nm). To address unwanted oscillations observed in longer tests (>2h), Stage 3 was introduced to enable the stable and consistent replication of activities at a physiological test rate. Despite higher RMSEs the axial error was 85.5 ± 24.27 N (equivalent to âˆ¼ 0.16 MPa), with shear RMSEs similar to other testing systems conducting pure moment tests at slower rates. The LIKE protocol enables the replication of physiological loads, providing opportunities for enhanced investigations of IVD mechanobiology, and the pre-clinical evaluation of IVD devices and therapies.