An Ultrasonic Elastography Method Based on Variable Length of Filter in Strain Computation

The elasticity of a biological tissue can be non-invasively positioned and quantified by ultrasonic elastography, which has a promising application prospect in medical imaging with indistinct boundaries for lesions, especially thermal ablation. This study aims to offer an ultrasonic elastography method that is based on a variable filter length. Compared with the conventional fixed filter length method and according to a strain gradient, the adaptive change of the filter length is allowed, and the motion boundary information between an inclusion and a background tissue is enhanced. With elastic phantoms and fresh porcine liver, the proposed novel method is estimated. The results show that better strain image presentation effects, clearer boundaries, and higher contrast-to-noise ratios (CNRs) have been achieved in regions of interest compared with the conventional method. The CNRs for in vitro experiment are 45.6, 95.0 and 141.2 of the method with M = 80, M = 290, and the proposed method, respectively. It is concluded that ultrasonic elastography based on variable filter length has great potential to be used in clinical applications. Future work will focus on more clinical studies on this method.