Automatic Computation of Potential Tumor Regions in Cancer Detection

Master Specialization: Visualization, Virtual Reality and Graphic Interaction
Topic Aproval date: 18/02/09
Orientation: research

Student: Bernat Orellana
Thesis advisor(s): Pere Brunet, Dr. Jorge Olsina, from the Vall d'Hebron Hospital in Barcelona, will participate in the project as an external advisor.
Contact: pere at lsi.upc.edu

The use of medical imaging and combined PET-CT data is becoming more and more important for efficient tumor detection (see references). The goal of this thesis is to develop new techniques for tumor finding, based on the computation of specific scalar fields that present maximum values at the regions having the highest probability of having tumors. The work will be structured in the following steps:

1. Analysis of the state of the art and discussion of the present techniques
2. Computation of a scalar field based on the local fractal dimension of volume isosurfaces.
3. Computation of a scalar field based on the local density of blood vessels
4. Analysis of the behaviour of these two algorithms when applied to real patient data.
5. Discussion and conclusions