Improving precision of tissue shear modulus quantification within theregion of acoustic radiation force excitation with compounded displacementestimates
The time-to-peak (TTP) displacement within the region of acoustic radiationforce excitation (ROE) is directly related to the tissue shear modulus. It haspreviously been shown that the ROE TTP can be directly converted to shearmodulus using a look-up table (LUT) for a given radiation force excitation focalconfiguration. While this method has the advantages of requiring less inputdata for stiffness reconstruction, ease of experimental implementation, and highspatial resolution, it suffers from lack of precision due to jitter inultrasonically tracked TTP displacement estimates. To reduce the variance of TTPmeasurements, compounding of TTP values obtained using different receiveapertures was investigated. A previously validated 3D FEM model was used tocalculate tissue displacement from impulsive radiation force excitation usingthe Siemens CH4-1 transducer and a fixed focal configuration (focal depth 49 mm,f/2, 2.2 MHz). Tissue was modeled as a linear elastic isotropic material withshear modulus from 1.5-20 kPa. Ultrasonic tracking of this displacement field inthe ROE was simulated in FIELD II using a receive spatial compoundingbeamforming technique. A single f/2 transmit aperture using 52 elements with afocal depth of 49 mm and 3.1 MHz center frequency was used. A total of 96elements were used for receive. These were grouped to form 9 overlapping receiveapertures each of 32 elements. The magnitude of TTP jitter after averaging TTPvalues from all 9 receive apertures was 0.250.02 ms for a simulated shearmodulus of 1.5 kPa and 0.12 mm displacement tracking kernel. In comparison, theTTP jitter without compounding using all 96 elements in a single receiveaperture was 0.560.07 ms. By using these compounded TTP values, the average ROETTP stiffness reconstruction error was 0.3 kPa, compared to 1.0 kPa withoutcompounding. Compounding TTP values from the different receive aperturesachieved TTP jitter reduction despite the fact that 1) speckle observed by thedifferent apertures is not fully decorrelated, and 2) receive aperture locationsoffset from the ROE axis have inherently higher TTP jitter. Moreover, thistechnique does not compromise spatial resolution, and can be experimentallyimplemented on a scanner capable of parallel beamforming without reduction ofpulse repetition frequency (PRF). © 2010 IEEE.
