Algorithm to estimate the level of elasticity of biological tissue with ultrasound elastography images

Abstract

The paper presents an algorithm to evaluate the elasticity of biological tissues from an ultrasound acquired elastography images to enhance the diagnostic capabilities. The algorithm was applied to RGB images of four healthy subjects; including pre-processing, filtering, region of interest selection, segmentation and pixel computing stages. A Repeatability and Reproducibility method (R&R) based on average and range technique is used to validate the procedure. For this test, two physicians acquired the US elastography images from the same subjects in the same anatomic region, at the same time. An elastometry graph is obtained after interpolation of the three color maps is applied. According R&R guidelines, the results for repeatability and reproducibility for each plane (red, green and blue) were less to 10%. These results are indicative of an algorithm with good performance. This proposal represents a promising diagnostic tool to evaluate objectively the elasticity of biological tissue for diagnostic, treatment follow-up and research purposes.