The spine is a complex and a still partially unexplored structure. Due to the constantly increasing longevity and the sedentary life-style typical of the industrialized countries, spinal diseases are becoming a serious problem inside the modern society and a deeper knowledge of spine biomechanics is required. In the literature a series of works investigated the biomechanics of the spine in order to characterize the overall spine, its organs, its tissues separately, in physiological, pathological and after treatment conditions. In vivo tests are performed evaluating the range or motion and the spinal loads in living subjects. In vitro tests allow measuring the strain on vertebrae and intervertebral discs and testing new spinal devices. Moreover, in silico tests are useful to simulate different loading scenarios, pathologies, and devices, especially when used in synergies with in vivo and in vitro tests. A methodology to merge the evaluation of the range of motion with the full-field strain measurement was not been implemented. A clear description of the strain associated to physiological tasks can help the clinicians, can improve the design of new devices and can guide the development of new surgical procedures. In order to achieve this goal, the mapping of the strain distribution must be achieved simultaneously on the vertebrae and the intervertebral discs. The aim of my PhD thesis was the implementation and improvement of the methodologies to quantify the displacements and strains in spine segments in a full-field view and a contactless way, in order to characterize the biomechanics of the spine as a whole and in the details of its organs.

The first part of this thesis focuses on the evaluation of the displacements and strains distribution on the entire spine surface. In order to achieve this goal a contactless, full-field measurement technique was employed and optimized:​ the Digital Image Correlation (DIC). Before starting to use the Digital Image Correlation, the tool was deeply validated, starting from mechanical specimen (aluminum beam) up to biological specimen (vertebrae). A factorial-design allowed optimizing the procedure to prepare a repeatable and reproducible speckle pattern and identifying the best acquisition/elaboration parameters. The optimization reduced the systematic error to 10 microstrain and the random error to 110 microstrain. Porcine spine segments, with intervertebral discs, were used to explore the feasibility of measuring strain on the entire specimen surface using the Digital Image Correlation. To measure the strain distribution on the hard and soft tissues, the spine segments were prepared with a random white-on-black speckle pattern and tested in simplified loading conditions to reproduce the anterior bending and the lateral bending. The displacements and the strains were thus evaluated, simultaneously on the vertebrae and on the intervertebral discs, using the optimized Digital Image Correlation showing the potentiality of exploring the spine as a whole and in detail.

The second area of research consists in going beyond the evaluation of measurements on the specimen surface, and providing the three-dimensional displacements and strains maps inside the specimen through Digital Volume Correlation (DVC). Of course the measurement uncertainties cannot be taken for granted. Even more than in the Digital Image Correlation, the reliability of Digital Volume Correlation must be assessed;​ because of no other measurement techniques are able to provide comparable measurements. Different studies were developed from the tissue-level (bovine cortical and trabecular bone cores) up to organ-level (porcine vertebrae and murine tibiae). Both laboratory microCT (voxel size 10-40 micrometers), and synchrotron radiation micro-CT (voxel size 1.6 micrometers – Diamond Light Source) were used to assess the impact of the quality and resolution of the input images. As no alternative measurement technique can be used to quantify the errors of DVC, multi-factorial (tissue types, imaging techniques, spatial resolutions, DVC approaches, DVC parameters) studies were designed based on datasets shared between different research centres (like Round-Robin test). This allowed to assess the effect of the single parameters on the final displacement and strain measurements. All the specimens were scanned twice without any repositioning and without any loads, because this procedure is the only way to know certainly the strains (zerostrain) within the specimen. Furthermore the algorithms were verified using artificially translated images. All these tests did not complete the validation of the Digital Volume Correlation that is still challenging, but defined the minimum and unavoidable intrinsic measurement errors related to the compromise between measurement uncertainties and measurement spatial resolution. The original goal was to obtain a measurement uncertainty lower than 200 micrometers in order to use the DVC also for the measurement of strain related to physiological loads. The threshold was reached with a measurement spatial resolution of ≈2mm for laboratory source microCT based DVC and 40/80 micrometers for Synchrotron radiation microCT based DVC. The acquired background about the optimization of the DVC parameters was exploited to start measuring the strain distribution within a vertebral body under load. This final work showed the strain gradients inside the vertebra in a destructive stepwise loading, highlighting, already in the elastic regime, the highest strain region where failure will start.

In conclusion, the project highlighted the importance of a careful validation before using these novel measurement techniques and confirmed that after optimizing the experimental details it is possible to apply these new procedures on spine segments. The methodologies can be considered as completed, but in the next years the application of the methods should be performed on human specimens:​ applying more complex loading scenarios and exploring the biomechanics in physiological, pathological and instrumented specimens.