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

INTRODUCTION

Finite Element (FE) analyses are crucial for studying biomechanics in human tissues and organs. Automating patient-specific (PS) Intervertebral Disc (IVD) models is challenging, particularly for investigating poromechanical behaviors under large deformations [1]. No comprehensive studies have examined how IVD morphology might contribute to disc degeneration (DD) [2]. This research aims to create a FE mesh repository of PS models, analyzing the relationship between IVD morphology and biomechanics using machine learning regression models.

METHODS

We utilized 169 3D lumbar IVD shapes from the MySpine project (FP7-269909), covering a range from healthy to Pfirrmann grade 4 degeneration, as obtained from MRIs [3]. The Bayesian Coherent Point Drift (BCPD) algorithm [4] aligned a validated structural FE IVD mesh [1] to these PS models in a seven-stage morphing process (Figure). We conducted mesh quality analyses and Hausdorff distance measurements, along with mechanical simulations under varying pressures, to simulate sleep and daytime activities, respectively, applying 0.11 and 0.54 MPa of pressure on the upper cartilage endplate (CEP). Simulation results were extracted from the anterior (ATZ) and posterior regions (PTZ) and the center of the nucleus pulposus (CNP). Data mining involved Linear Regression, Support Vector Machine, and eXtreme Gradient Boosting, focusing on mechanical variables relevant to DD. The modeling approach was validated using experimental data from [2], including FE-predicted displacements under a 500 N compressive load.

RESULTS

A 94% similarity was obtained between the AF and NP segmentations with their respective meshes. The mechanical variables, such as pore fluid velocity, show substantial variations in the transition zone (TZ) between AF and NP. Clinical imaging has shown DD onsets in the same TZ [5]. The relative error between the simulation of this validation and the experiment was 5.20% and presented a similar curve to the simulation of the previous work. Local morphological variables significantly impacted the local mechanical response. The local disc heights, respectively, in the mid (MH), anterior (AH), and posterior (PH) regions, were key factors in general.
Additionally, mechanical responses in the PTZ, CNP, and ATZ were influenced by the shape of the upper and lower CEPs.

DISCUSSION

The cohort is a unique set of models to explore the effect of multiple geometric variations on the multiphysics and mechanobiology of IVD, including the full organ tissue structure [2]. The algorithm can create PS FE models from any segmented surface provided by third parties and generate representative synthetic models of healthy and DDs. This study suggests that disc morphology affects Mechanical variables of interest in DD. Attention should be paid to the antero-posterior distribution and local effects of disc heights. Surprisingly, the CEP morphology remotely affected the fluid transport in NP volumes around mid-height, and mechanobiological implications shall be explored. In summary, PS IVD modeling is poised to reveal crucial associations between IVD phenotypes and local tissue behaviors, contributing significantly to the understanding of IVD mechanics and pathophysiology.

ACKNOWLEDGMENTS
European Commission: Disc4All-MSCA-2020-ITN-ETN GA: 955735 O-Health-ERC-CoG-2021-101044828