Strain Elastography & Tissue Characterisation as a Tool to Differentiate Tumor from Healthy Tissue

Abstract

The purpose of this study was to compare diagnostic values of normal and effected tissues with two techniques using strain elastography and tissues characterization. This study was carried out on a breast phantom containing all human body parameters. Analysis was performed using a lone phantom to correlate a relation between the values of Strain Elastography (SE) and first order texture parameters results. For SE SonixTouch Q+ (Ultrasonix Medical Corporation, 130-4311 Viking Way, Richmond, Canada) device using a linear-array ultrasound probe at a frequency of 10MHz with a gain of 40%. Elastography breast phantom was purchased from CAE healthcare USA, 3600 Edgelake Drive Sarasota FL, USA. For tissue characterization a Region of Interest (ROI) that encompasses both (normal and stiffer) areas were selected. MAZDA software was used to carry out the image analysis (mean and variance) of the tumour and healthy tissue, ROI of 1600 pixels at both regions was selected. An affirmative and resilient outcome was observed between the numerals of normal and tumor tissues, both for SE and first order texture parameters values. After our study we suggest that SE and tissue characterisation via first order texture parameter is a reliable technique to highlight normal and tumor tissue (with respect to same reference, for SE technique only). SE and first order texture parameters (mean and variance) paved way in highlighting the breast tumors fully. It is suggested that SE being more reliable approach in determining the stiffness for breast lesion, as it produces the results with real time imaging. However texture parameter gives an objective assessment of the image with a discriminating feature of the tissue.