A whole spine multibody model for predicting injury in high-rate axial loading
File(s)
Author(s)
Low, Lucas
Type
Thesis
Abstract
Explosives have become one of the largest contributors to injury in conflicts over the past two decades. When explosives are detonated under a vehicle – termed underbody blast – the load is transferred to the occupant through the pelvis into the spine, resulting in high-rate axial loading which causes spinal injury. This PhD aims to develop a framework for generating patient-specific multibody whole spine models to study the kinematics and kinetics of the spine in underbody blast.
A methodology for generating patient-specific multibody models from computed tomography scans was developed. A literature review found scarcity of data on the structural properties of the spine in underbody-blast-relevant high-rate axial loading. Therefore, experiments were conducted on intervertebral discs from four whole cadaveric human spines, across three orders of magnitude of compressive strain rates and flexion rotation rates. An inverse multibody model of the experiment was developed to analyse the experimental results.
With updated structural properties determined from experiments, the multibody spine model was calibrated against lab-simulated whole body cadaveric underbody blast tests. An optimisation algorithm was implemented to aid in calibration. Three patient-specific spine models were developed and calibrated against their respective underbody blast tests.
The three patient-specific models were repositioned from the baseline posture to two additional postures – leaned back and slouched forwards. The models simulated a range of underbody-blast-relevant loadcases, and injury risk was predicted using injury risk curves from the literature. Increased body mass index was found to increase injury risk in the lumbar spine; the leaned back and slouched forwards postures presented reduced injury risk in the cervical spine.
Future work on the spine model includes the validation of the subject-specific models, and development of additional anthropometries to capture the variation in the population. Additionally, implementing a seat will allow the model to aid in designing and assessing injury-mitigation technologies.
A methodology for generating patient-specific multibody models from computed tomography scans was developed. A literature review found scarcity of data on the structural properties of the spine in underbody-blast-relevant high-rate axial loading. Therefore, experiments were conducted on intervertebral discs from four whole cadaveric human spines, across three orders of magnitude of compressive strain rates and flexion rotation rates. An inverse multibody model of the experiment was developed to analyse the experimental results.
With updated structural properties determined from experiments, the multibody spine model was calibrated against lab-simulated whole body cadaveric underbody blast tests. An optimisation algorithm was implemented to aid in calibration. Three patient-specific spine models were developed and calibrated against their respective underbody blast tests.
The three patient-specific models were repositioned from the baseline posture to two additional postures – leaned back and slouched forwards. The models simulated a range of underbody-blast-relevant loadcases, and injury risk was predicted using injury risk curves from the literature. Increased body mass index was found to increase injury risk in the lumbar spine; the leaned back and slouched forwards postures presented reduced injury risk in the cervical spine.
Future work on the spine model includes the validation of the subject-specific models, and development of additional anthropometries to capture the variation in the population. Additionally, implementing a seat will allow the model to aid in designing and assessing injury-mitigation technologies.
Version
Open Access
Date Issued
2023-11
Date Awarded
2024-06
Copyright Statement
Creative Commons Attribution NonCommercial NoDerivatives Licence
Advisor
Masouros, Spyridon
Newell, Nicolas
Sponsor
Royal British Legion
Publisher Department
Bioengineering
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)