The past decades have seen the introduction of a variety of medical imaging modalities such as as magnetic resonance imaging (MRI), computed tomography (CT), positron emission tomography (PET) and ultrasonography. In recent years this development has accelerated and a variety of new techniques for measuring tissue properties, fiber orientation and functional processes have been proposed. Consequently the available medical data sets now comprise measurements ranging from scalar fields such as tissue density (x-ray) and water content (MRI) to vector fields, such as blood flow velocity, and tensor fields such as myocardial strain and cellular water diffusion.
The size and complexity of medical data available today makes it difficult to analyze and to understand them. This is particularly true for multi-dimensional data such as vector and tensor fields, and for the simultaneous analysis of multiple data sets such as MRI, PET and CT. The understanding of the data can be improved by visualizing it.
We have developed a toolkit for visualizing biomedical data sets and finite element models which are becoming increasingly important for understanding and simulating organ function and diseases. Using the finite element data structure allows the definition of tissue properties in material coordinates, enables the selection of important structural components of the modeled organ (such as the inside or outside surface of the heart) and facilitates the computation of performance measures.
An expression defining a field can also contain conditional statements that can be used to create a segmentation function for different tissue types. Using this tool we can create the visualization shown in part (a) of the figure below. Regions of gray matter, white matter, and Celebral Spinal Fluid are coloured blue, red and green, respectively. Part (b) of the figure shows a slice visualized with a barycentric colour map which uses three interactively derived fields which are positive and add up to one for all points.
We would like to thank Peter J. Basser from the National Institute of Health, Bethesda, MD, for valuable discussions and Carlo Pierpaoli for providing us with the diffusion tensor data set of a healthy brain used in some of the above images. We would also like to thank Alistair A. Young from the Department of Physiology abd the Department of Anatomy with Radiology of the University of Auckland, Auckland, New Zealand, for valuable discussions and for providing us with the models of the left ventricle. Finally our thanks goes to Dr. Richard White of the Cleveland Clinic, Cleveland, Ohio, USA, who kindly provided tagged MRI data of a left ventricle diagnosed with dilated cardiomyopathy.
Modified: 01/22/2003 15:23:58 by Burkhard Wünsche (Wuensche)