Here is a snippet of the page about my thesis:
Classification for Volume Rendering in
My research topic was in the exciting field of semi-automatic classification of medical volume data. The challenge is, that there is no trivial way of classifying medical volume data. Our novel approach was to use the tools and advantages of Augmented Reality. A image guided system with infrared light based tool tracking of the company BrainLAB allowed us to use tools to choose regions of the medical volume data set and assign it a colour. Neural networks were trained on a set of such data and applied to the entire dataset. That way, we achieved an intuitive way of how to classify a medical volume data.