Elasticity estimation based on ultrasound medical images

• Grant number: 10/18285-3
• Support Opportunities: Regular Research Grants
• Start date: June 01, 2011
• End date: November 30, 2013
• Field of knowledge: Engineering - Biomedical Engineering - Medical Engineering
• Principal Investigator: Sergio Shiguemi Furuie
• Grantee: Sergio Shiguemi Furuie
• Host Institution: Escola Politécnica (EP). Universidade de São Paulo (USP). São Paulo , SP, Brazil

Abstract

Medical images provide information about the interaction of physical phenomena, such as x-ray, with tissues and organs. From these interactions, the aim is to infer the composition of the material, as well as physical and physiological properties of cells, tissues and structures. An important feature, from the clinical point of view, is the rigidity of part of an organ, for example, of the kidneys, traditionally investigated through palpation. On the other hand, the low stiffness is indicative of the vulnerability of coronary lesions. In this project we investigate the elasticity of structures (elastography) and their components through dynamic ultrasound images. To do this, we need a temporal sequence of images from the same region at different pressures. We will use ultrasound, because it is a modality of relative low cost, high temporal resolution, and high availability. We intend to explore the richness of spatial and temporal information present in conventional images to characterize them objectively. The proposed methodology consists of the following: a) detailed numerical simulations of 2D and 3D dynamic images with deformation and realistic image noise; b) estimate of displacements and deformation based only on images; c) estimate of the elasticity. Conventional elastography is based on images obtained by sampling echo signals at high spatial resolution so as to preserve the oscillatory characteristics of RF signals, implying the need of specific hardware. We intend to investigate elastography using conventional images based on envelope. We are proposing new approaches for displacement and strain estimation based on signature tracking. The assessment of the proposed methodology will be carried out in two phases. First, by means of numerical simulation, in which we will create structures and lesions with contractions. The formation of images will incorporate realistic noise with speckles. The results obtained by the proposed approach will be compared with the expected ones. In the second phase, we will use real images in vitro. Innovations and potential contributions are focused on: a) environment to explore dynamic ultrasound images on conventional and RF modes; b) alternative proposals for the estimation of deformations and displacements; c) quantifying and classifying lesions in tissues and organs. (AU)