A Feature-Based Morphing Methodology for In-Vivo Strain Assessment in Biological Structures

It is often important, for diagnostic purposes, to evaluate quantitatively the motion undergone by a biological structure, starting from a viable tomographic imaging technique, such as Computed Tomography (CT) or Magnetic Resonance (MR). This is often associated with the need of the clinical personnel to perform an evaluation of a risk factor associated with pathology, e.g. ventricle performance [1]. As of today, the problem has been approached by using the magnetic resonance (MR) tagging technique [2] or experimentally by determining the location of a series of surgically implanted markers in each temporal phase [3]. The latter procedure is naturally very invasive, while MR tagging requires the use of sophisticated MR sequences that are rarely employed in standard care. Multi-detector row computer tomography (CT) scanners (routinely used in cardiac pathology diagnosis) as well as routine cine-MR sequences, on the other hand, allow dynamic imaging of the heart and large vessels with cardiac gating. In order to exploit these dynamic sequences, we have developed a tag-less method to estimate local strains from dynamic tomography gated images.