Medical Applications of Dual Energy Computed Tomography
PR/BA/DA
Description
Dual Energy Computed Tomography (DECT) is a recent data acquisition modality in the medical domain. A patient will be scanned at two different energy levels in contrast to the common Computed Tomography (CT), where only one energy level is applied. On the one hand the radiation dosage for the patient is increased, but on the other hand, more data is generated that might provide advantages. In DECT, the two obtained data sets feature different properties concerning noise and contrast of different types of tissues. For example, the lower energy data set is noisier but it is easier to distinguish the various types of tissues, whereas the higher energy level produces less noise data sets but with less contrast. One possibility is to denoise the low energy volume using bilateral filtering with the high energy level defining the weight. The goal is to utilize and combine these properties to segment, for example, bones and vessels, or even one step further, identify soft plaque or calcifications, which are reasons for pathologies such as vessel occlusions.
Tasks
Your task is to study and visually evaluate the usefulness and applicability of DECT for medical purposes. You will work within a software environment that is used in the daily clinical routine of two hospitals and collaborate with medical doctors. You will use modern graphics technology such as CUDA and parallel programming on the CPU. Although you will work along a defined way, we encourage you to contribute even with your own ideas.
Requirements
- Basic knowledge in computer graphics and visualization
- Programming in C++
- Knowledge in parallel programming (CUDA/TBB) is advantageous
- Knowledge of cmake and Qt is advantageous
Environment
The project should be implemented as a plugin in the AngioVis Framework. Under the following link, http://www.angiovis.org/, you can find additional information concerning the framework and its application areas.
Contact
Email: gabriel.mistelbauer { at } cg.tuwien.ac.at
Homepage: Gabriel Mistelbauer
