Filtering and Visualization of Scalar Fields on the GPU

OData support
Dr. Szirmay-Kalos László
Department of Control Engineering and Information Technology

Wilhelm Conrad Röntgen showed the world the very first X-ray picture in January of 1896, it presented the hand of his wife. The finger bones and the silhouette of a ringer could have been seen. The importance of the discovery was quickly recognized and during the early years of the twentieth century hundreds of medical stations based on radiography have begun operation.

Since then medical imaging is vital part of the diagnostics. With the help of X-rays, good radiograms could be taken from the skeletal system. Also it's capable of the detection of soft tissue diseases. While X-ray is producing 2D radiographs, CT is capable of creating 3D volumes. MRI is for the same purpose but based on completely different technology. It can detect the distribution of water inside the body and thus provides a good way to distinguish soft tissues.

Conclusion can be drawn from the above that even nowadays medical imaging is advancing. New techniques provide different ways to peek inside the body. Nevertheless the visualization of the volume takes effort and also the source can be noisy.

The thesis provides solution to filter and visualize in real time the raw data, all this on ordinary graphics cards. As we are talking about data with the extent of millions of bytes and as it must be processed with non linear algorithms, the best way to speed up the computation is to do it parallel. This is supported by NVIDIA's CUDA technology which I'm going to use through the thesis.

In the visualization my goal was to provide valuable information for diagnostics and at the same time be aesthetically pleasing. To achieve this I used OpenGL and it's shader language GLSL.


Please sign in to download the files of this thesis.