Model tags: direct three-dimensional tracking of heart wall motion from tagged magnetic resonance images

AA Young - Medical image analysis, 1999 - Elsevier
Medical image analysis, 1999Elsevier
Although magnetic resonance tissue tagging is a useful tool for the non-invasive
measurement of three-dimensional (3-D) heart wall motion, the clinical utility of current
analysis techniques is limited by the prohibitively long time required for image analysis. A
method was therefore developed for the reconstruction of 3-D heart wall motion directly from
tagged magnetic resonance images, without prior identification of ventricular boundaries or
tag stripe locations. The method utilized a finite-element model to describe the shape and …
Although magnetic resonance tissue tagging is a useful tool for the non-invasive measurement of three-dimensional (3-D) heart wall motion, the clinical utility of current analysis techniques is limited by the prohibitively long time required for image analysis. A method was therefore developed for the reconstruction of 3-D heart wall motion directly from tagged magnetic resonance images, without prior identification of ventricular boundaries or tag stripe locations. The method utilized a finite-element model to describe the shape and motion of the heart. Initially, the model geometry was determined at the time of tag creation by fitting a small number of guide points which were placed interactively on the images. Model tags were then created within the model as material surfaces which defined the location of the magnetic tags. An objective function was derived to measure the degree of match between the model tags and the image stripes. The objective was minimized by allowing the model to deform directly under the influence of the images, utilizing an efficient method for calculating image-derived motion constraints. The model deformation could also be manipulated interactively by guide points. Experiments were performed using clinical images of a normal volunteer, as well as simulated images in which the true motion was specified. The root-mean-squared errors between the known and calculated displacement and strain for the simulated images were similar to those obtained using previous stripe-tracking and model-fitting methods. A significant improvement in analysis time was obtained for the normal volunteer and further improvements may allow the method to be applied in a ‘real-time’ clinical environment.
Elsevier