doi:10.1016/j.ultrasmedbio.2005.09.015 ● Original Contribution IDENTIFYING THE MECHANICAL PROPERTIES OF TISSUE BY ULTRASOUND STRAIN IMAGING EMRE TURGAY,SEPTIMIU SALCUDEAN, and ROBERT ROHLING Department of Electrical and Computer Engineering, University of British Columbia, Vancouver, BC, Canada (Received 5 April 2005, revised 2 September 2005, in final form 14 September 2005) Abstract—A new elastography method is presented that images the mechanical properties of soft tissue. The tissue is externally vibrated over a range of frequencies simultaneously and the resulting displacement is recorded at multiple locations and time instants with a sequence of ultrasound images. Two methods are proposed for estimating the mechanical properties from the recorded data. In the first method, equations of motion are written for a tissue model of masses, springs and dampers. The equations are solved to identify the model parameters and construct an image. The second method performs a transfer function analysis of the tissue motion. The tissue properties are identified from the shape the of transfer function. Simulations show good performance of the new method compared with static elastography. Initial experimental results from homogeneous and layered tissue phantoms also demonstrate the ability quantitatively to image tissue stiffness. Preliminary results are obtained for viscosity and density estimation. Further work is needed to extend the formulation to 3-D and improve robustness of the viscosity and density estimates. (E-mail: tims@ece.ubc.ca) © 2006 World Federation for Ultrasound in Medicine & Biology. Key Words: Elastography, Stiffness, Viscosity, Density, Transfer function, Mass-spring-damper model. INTRODUCTION The mechanical properties of tissue depend on many factors, including the type of tissue (e.g., fat, muscle, blood) and the presence of disease or pathology. It has been known that many cancers, such as carcinoma of the breast and the prostate, appear as hard nodules (Ophir et al. 1999). Manual palpation is widely used for detecting these nodules, but detection by palpation is restricted to large tumors that reside relatively close to an accessible surface (Sumi et al. 1995). Current imaging devices, such as computed tomography, magnetic resonance imaging (MRI) and ultrasound (US) are not directly capable of measuring the mechanical properties of tissue. Imaging the mechanical properties of tissue has become the sub- ject of increasing interest during the past two decades. The techniques used for imaging these properties of tissue form a new field of study called “elastography” in the literature. All elastography techniques combine a form of mechanical excitation with measurement of the resulting tissue motion. One group of elastography tech- niques uses quasistatic (below 10 Hz) compression ap- plied externally to the surface of the tissue (Sumi et al. 1995; Ophir et al. 1991; O’Donnel et al. 1994; Sko- voroda et al. 1995). The resultant tissue deformation is then related to local stiffness of the region. However, the static elastography techniques only measure the static behavior of the tissue and not the dynamics. This means that stiffness is usually estimated, but not viscosity, den- sity or other mechanical properties. In another approach, focused US is used to induce localized tissue motion. Tissue deformation is then analyzed to obtain local tissue properties (Nightingale et al. 2001; Viola and Walker 2001; Fatemi and Greenleaf 2000). However, specialized high-intensity US beams are needed and the process must be repeated at many locations to obtain a full image. In a third approach, an external vibration source (above 10 Hz) creates shear waves in the tissue. The mechanical properties are then extracted by analyzing the wave propagation and decay (Lerner et al. 1988; Yamakoshi et al. 1990; Levinson et al. 1995; Sinkus et al. 2000; Sandrin et al. 1999; Kruse et al. 2000). Doppler imaging has also been used to measure tissue elasticity (Lerner et al. 1988; Yamakoshi et al. 1990; Levinson et al. 1995). However, these techniques are often limited by high measurement noise from the Doppler data. With the Doppler techniques, tissue viscosity has been introduced Address correspondence to: Dr. S. Salcudean, Dept. of Electrical and Computer Engineering, 2356 Main Mall, UBC, Vancouver, BC V6T 1Z4 Canada. E-mail: tims@ece.ubc.ca Ultrasound in Med. & Biol., Vol. 32, No. 2, pp. 221–235, 2006 Copyright © 2006 World Federation for Ultrasound in Medicine & Biology Printed in the USA. All rights reserved 0301-5629/06/$–see front matter 221