Towards non-invasive automatic detection of cardiac pathology by strain and rotation analysis

We propose a new MR tagging image analysis based on a first order approximation of the multi-scale optical flow constraint equation [2] applied to the tag phase information, to assess left ventricle contraction patterns, and derive parameters such as strain and rotation. We have tested our method to both phantom and in vivo volunteer and patient data. The new optical flow based motion analysis method is able to detect differences in both minimal strain E2 and amount of rotation in the basal slice between healthy volunteers and patients. The new method combines two perpendicularly encoded tagging series, and does not suffer from tag fading. Calculation of high-resolution strain and LV rotation from the first-order optical flow solution is straightforward. The method is algorithmically very elegant in that it requires only off-the-shelf algorithms