Honours Projects
Contact: Dr Nicole Bordes (nb@maths.uq.edu.au - tel: (07) 3365 7506)
This list is under perpetual construction: more projects will be added later during the year.
| Analysis of intensive care unit medical data (maths/science students) |
| A series of physiological data have been acquired over time. Several of the variables are inter-related. A doctor will look at a "high level" variable and if it is abnormal will look at the inter-related variables that could shed some light on why the high level variable is abnormal. (for instance ph is the high level variable, and the other related variables are levels of paCO2 and HCO3). This project aims to use statistical and visualisation techniques to model physiological data so that it can be represented in ways that are physiologically consistent and psychologically relevant to decision makers. |
| Extraction of blood vessels in liver for surgery planning (maths/science students) |
|
One of the treatments of liver cancer is resection or ablation of one or
several tumours and of an area of healthy tissue around it. The extraction
of the essential information from Computed Tomography (CT) scans is time-consuming:
the radiologist must trace the contour of the liver manually as well as the
tumour(s) and the main vessels. In addition blood vessels and liver tissue
show similar contrast on the CT scans. This project aims to use image
processing procedures to extract automatically the blood vessels from
the CT scans of the liver.
  Figure 1: a) view of the major blood vessels with respect to the tumour. b) view of the major blood vessels with respect to the tumour and the liver. |
| Visualisation of large multi-resolution images (ITEE or maths/science students) |
|
Visualisation of multi-resolution images such as those obtained in
microscopy or astronomy is a challenging problem. One method used
to visualise images capturing light intensities is to create height
fields however in some cases large meshes are generated resulting
in high use of memory and processing. Displacement shading is a technique used in computer graphics: a displacement shader displaces each point on a surface by a small amount, before the rendering is performed. This offers significant performance advantages over creating and manipulating a large mesh. We propose to develop a displacement shading algorithm to visualise such images and to compare this technique with traditional height fields visualisations.
Using different level of detail the student will implement an algorithm (possibly bump mapping) to display data acquired at different resolutions by using different levels of detail.
|
| Spatial analysis of Great Barrier Reef geographical data |
|
Under construction
|
| Access Grid: application sharing across multiple sites (ITEE students) |
|
Under construction
|
| Tiled display: chromium (ITEE students) |
|
Under construction
|