Knowledge and Semantics Encoding in Medical Volume Rendering

Todays pletora of state-of-the-art medical imaging modalities produce a wide selection of data types with different requirements and purposes, a diversity that cannot be handled by traditional viewing systems. This project aims to exploit the visualization user's domain knowledge in new processing and data management techniques that enables efficient human anlysis of medical data sets from different modalities in a clinical environment.


The population pyramid and increased life span in the western world will lead to greatly increased need of health care. The society needs to make health care much more efficient to cope with this financial burdon. The field of medical imaging is currently facing a great challenge, the data explosion. During the 20th century, the focus was on retrieving as much medical image data as possible. Today, the task is to navigate within an abundance of data, to pinpoint and highlight the diagnostically important features while discarding irrelevant data. The technical development is not rapid enough to match the constantly increasing number of modalities. This is, and will be, a severe bottleneck for the medical workflow that needs to be handled.

Project aim

The project aims at improving interaction, design and quality of the medical workflow with focus on semantic descriptions and applicability. Topics include knowledge assisted methods, multi-modal techniques and scalable algorithms for improving the state-of-the-art in Direct Volume Rendering and dual-energy CT classification, virtual autopsy and large volume visualization in particular.

Project Members