Analysis of the impact on the quality of an elastogram, of the algorithms for strains calculation

This work makes a study about obtaining elastography images, or elastogram from ultrasound RF frames. This technique is very useful in the improvement of medical diagnosis of diseases affecting the biological tissues, including cancer. One of the decisive factors for the quality of a clinical diagnosis is the precision of the results derived from a clinical test. In particular, in this work the quality of an elastogram is researched, as a result of the robustness of the algorithms to obtain the coefficients of axial and lateral deformations (strains), which are the basis for an elastogram. A comparison is made among three algorithms reported in the literature to obtain strains, namely, exhaustive block search, a hybrid multilevel search method based on cross correlation for real-time, and a not based on blocks method that uses dynamic programming approach. As results of this work, a software application that allows to generate elastograms for research purposes has been built, and a preliminary analysis showing that new algorithms for strains computing must be proposed in order to reconcile the needing of real-time calculation with the needing of robustness of the information provided by the elastogram. Some guides for this end can be extracted from the obtained results.