2D and 3D high-spatial and high-temporal resolution ultrasound imaging for characterization of tendon mechanics.

Achilles tendinopathy affects competitive and recreational athletes as well as inactive people. The incidence of this pathology in the general population is 1.8%and over 80% of tendon ruptures occur during recreational sports. The etiology of tendinopathies is multifactorial, but tendon overuse is assumed to be one of the main pathological stimuli.

Recently published studies have demonstrated the need for a better characterization of the tendon, more specifically, the local tendon mechanics. By performing this characterization in an accurate and reproducible manner, it isexpected to understand better the mechanical behaviour of healthy tendons and, consequently, better understand the tendinopathy pathogenesis.

Nowadays, characterization of tendon mechanics is done by estimating the global tendon strain and the local tendon tissue displacement. Conventional ultrasonography (6-10MHz) is the imaging modality used for this characterization due to its ubiquitous application for tendinopathy diagnosis and easy accessibility. However, it is believed that conventional ultrasonography is limited due to its low spatial and low temporal image resolution. Furthermore, the high-variability of tendon strain estimations reported in the literature are also assumed to be dependent on the imaging system used for the characterization of the tendon mechanics.

This work presented the investigation of the implementation of a 2D highspatial and high-temporal resolution ultrasound acquisition system for the in-vivo characterization of tendon mechanics. Due to the novelty of this type of acquisition system for tendon strain estimations, a validation step was initially designed. For this validation, 2D global strain estimations were calculated using a developed affine image registration method. Good performance was obtained for this validation step, which allowed the validation of the image registration method as well as the validation of the novel high-spatial and high-temporal resolution ultrasound acquisition system. Afterwards, a 2D deformable image registration method was developed for the in-vivo characterization of the local tendon mechanics. Local tendon mechanics was evaluated by estimating regional tendon strain and local tendon tissue displacement. The developed deformable image registration method and the calculation of the features used for the characterization of the local tendon mechanics were embedded in the KULTeC. The KULTeC is a standalone, easy and intuitive application, which is intended to allow the clinician to easily characterized the local tendon mechanics. Due to the absence of ground-truth for the in-vivo data, a psychometric approach was performed. The obtained results confirmed the reproducibility of the implemented deformable image registration method, demonstrated a convergence of the estimated results with the literature results and demonstrated the capability of the method for the discrimination between asymptomatic and medium to severe tendinopathy conditions.

After the in-vivo characterization of the local tendon mechanics, the impact of out-of-plane motion was investigated. This type of artifact, which affects every 2D imaging modality, has frequently been pointed out as tendon strain estimation errors. For this investigation, 3D high-spatial resolution ultrasound data was used, and the global strain was estimated using an affine image registration approach. The results obtained from this experiment, using in-silico data, demonstrated the theoretically added value of strain estimations using 3D data. However, with the increasing complexity of in-vitro and ex-vivo data, and the increasing complexity of the acquisition setup, the acquisition of 3D high-spatial resolution US images becomes very challenging.

In conclusion, this work demonstrates the successful implementation of 2D high-spatial and high-temporal resolution ultrasonography system for in-vivo characterization of the local tendon mechanics. The high-spatial  and high-temporal resolution provided by the used acquisition system together with the developed deformable image registration  method have the potential of allowing better and more accurate in-vivo characterization of local tendon mechanics. Furthermore, the developed KULTeC application allows an easy and intuitive use of the developed image registration method in the clinical practice. The results obtained for 3D high-spatial resolution ultrasonography for global strain estimations show that this modality has the potential for elimination of out-of-plane motion but its application to real-world conditions is still technically challenging.