PDF version of this entire document


Analysis of the heart is a problem which is principally associated with biological factors that require knowledge of anatomy, but the problem can be simplified by pre-processing, alternative visualisation paradigms, and occasionally an automated classification of common groups. In this work, the problem gets narrowed down such that only a portion of the heart is dealt with. The key goal is to perform an analysis of human heart images while overcoming its different types of movements which blur tagging and make it increasingly difficult to study the anatomy.

We are attempting to exploit a novel approach whereby an isolated portion of the heart is captured for standalone analysis2 as the shape of this portion is changed, over time.

DICOM-formatted datasets are to be analysed using calculations of stress (e.g. strain rotation, principal strains, and maximum shear strain), which can be derived by isolating and segmenting the data along the different axes.

Roy Schestowitz 2010-12-25