Caracterización elastográfica de tejidos biológicos y su aplicación a la detección precoz del cáncer de mama

Laburpena

Elasticity imaging or elastography is the most recent modality of ultrasonic imaging, aimed to produce stiffness maps of tissue. Since malignant tumors are significantly stiffer than the surrounding healthy tissue, elastography is considered a highly specific imaging modality for cancer diagnosis. Many different ultrasound elastography variants have been developed from the pioneering works 25 years ago. In all of them the hardness is estimated from the displacements of tissue subjected to a stress field. The differences arise in the ways force is applied to produce stress and the methods used to record displacements. The constitutive equations of any material relate stresses with strains. In this work, tissue is considered a pure elastic medium, where stress and strain tensors are linearly linked without hysteresis. This leads to a simple relationship between the Young and the shear elastic moduli of tissue. Further, it is found a straightforward relation between the elastic moduli and the propagation velocity of shear waves. Displacements are evaluated by cross-correlation techniques applied to A-scans before and after the application of stress. The image speckle displacements are measured this way and estimated strain is displayed as a qualitative indication of stiffness. Static elastography produces the stress field by the (slight) pressure of the transducer in contact with the patient skin. Only qualitative elasticity maps can be produced since the elastic modulus cannot be obtained from unknown contour conditions. An alternative is provided by the Acoustic Radiation Force (ARF) generated by an ultrasonic beam focused at some depth into the tissue. Measured displacements in the focal region provide a qualitative estimation of hardness, as in static elastography. But the simultaneous production of shear waves, opens the possibility of numerically computing the local elastic modulus from its transversal propagation velocity. ARF techniques provide a means to produce localized displacements in some extent around the focus. To increase imaging range, multi-focal techniques that combine the results obtained with different focal depths are proposed. The technique is applied to sector images to also increase laterally the image range. However, a normalization procedure is required to decouple displacements variations due to elasticity changes from those caused by variations in beam intensity with depth, attenuation, focal distance and steering angle. A closed-formulae based procedure that does not require prior knowledge of the elastic properties of the medium is proposed. Furthermore, it is found that ARF can be applied to tissue without contact with a coupling medium (water) between the array and the tissue. This opens the opportunity of further increase the image lateral extension by compounding several side-by-side acquired sector elastographic images, which constitutes an important innovation. This leads to the possibility of performing automated acquisitions of Full Angle Spatial Compounded (FASC) elastographic images, which improves resolution, contrast and contrast-to-noise ratio, as it is demonstrated. Development of a methodology for elastography FASC, its implementation on an automated breast imaging system and its experimental validation are key contributions of this Ph. D. Thesis. Theoretical and experimental work support to the new elastographic techniques and methods proposed in this Ph. D. Thesis, which has been directed to improve the early detection and diagnosis of breast cancer