Résumé: The articulations of the horse of competition are very solicited, which may result in multiple tendinopathies such as tearing or rupture of the Superficial Digital Flexor Tendon (SDFT). Ultrasound imaging, inexpensive, non-invasive and conducive to explore the internal structure of the tendon is regularly used to assess the integrity of the SDFT. However, the functional prognosis on the state of the SDFTs from the clinical B-scan images is often difficult. This difficulty results from the lack of prior information of the observed structures on the B-scan images, which can lead to subjective interpretation or incorrect statement on the integrity of the tendon. The absence of imaging tools also limits veterinarians in their clinical decision. In this research, we focus on developing an analytical method to objectively assess the internal structure of the SDFTs from clinical B-scan images. Two axes of research are chosen: - A simulation, under realistic conditions, of the propagation of ultrasonic waves through the tendon structure in order to reproduce the clinical appearance of the clinical ultrasound B-scan images. The simulation results helped us to deduce information about the structural content observed in the clinical images of normal and injured SDFTs. This information was then used for: - A segmentation of the internal structure and quantification of the fiber fascicle bundle densities of the SDFTs from clinical ultrasound B-scan images. The simulation allowed us to note objectively that the hyperechoic structures observed on Bscans of normal SDFTs are the results of reflection and scattering of the ultrasonic waves at the interfascicular spaces that surround fiber bundles. The interfascicular spaces that generate hyperechoic structures at a frequency of 7.5 MHz are those which have thicknesses greater than the acoustic wavelength ? (~ 256 m) along the axis of the ultrasonic wave propagation. The segmentation method used can be summarized into two main approaches: the first approach is dedicated exclusively to the SDFT B-scans and it combines the logarithmic decompression and morphological operations. A second approach, more general, is used to segment SDFT B-scans and macro-photographic images. This later is based on a new morphological thinning algorithm. Quantification of the internal structure of SDFTs discriminated objectively normal from injured SDFTs. Statistical analysis on the fiber fascicle bundle densities over the preferential sites of SDFTs showed a low density at the distal site. This analysis was corroborated by applying the same segmentation and quantification approaches on a macro-photographic image dataset of SDFTs. This structural observation may reflect the weaknesses of the mechanical strength of the distal site and its predominance for injuries. In conclusion, these results could possibly be adapted to the study of human tendons and ligaments.

Edition: Canada: Universitè Quebec
Langue: Français
Collation: 213 p. ill. ;30 cm.
Diplôme: Doctorat d'état
Etablissement de soutenance: Universitè Quebec