ENGG7808 - Engineering Postgraduate Project D (#8)
Commencing Semester 2 2009
Coordinator: Adam Postula (adam@itee.uq.edu.au)
Project List
Industry projects may also be available - see the CEED website for details.
Marek Bialkowski
Office: 78-527
Phone: 53563
Email: meb@itee.uq.edu.au
1 - Microwave Imaging Radar System for Breast Cancer Detection
| Supervisor: | Marek Bialkowski | Project ID: | 1 |
|---|---|---|---|
| Research Group: | Electromagnetics and Imaging Group | Max. students: | 2 |
| Discipline(s): | Microwaves and Radar | Num. students signed up: |
0 |
| Description: | The project concers theoretical and experimental invetsigations into capabilities of an Ultra Wideband (UWB) Microwave Radar system to detect breast cancer. The project will be carried by a team of MEng and BE final year students and the tasks will concern the following issue: a. Design and development of a turntable and associated control software to provide an UWB probe antenna scan over a cylindrical surface. During this scan, reflection type measurements on the breast phantom will be performed with a vector network analyser (VNA) over a UW frequency band and the data will be transformed to the time domain using time domain capability of VNA. In addition, the work to be undertaken may include b. creating breast phantoms with electric properties similar to those of healthy breast tissue. c. Image formation and processing The collected experimental data will be used to form spatial images of the breast phantom. Suitable signal processing will be applied to compensate for the signal drop with distance to enhance detection of tumour targets just by visual inspection of the created image. d. CST Microwave Studio, FEKO or Ansoft HFSS simulations of breast radar The choice of specific tasks (a, b, c or d can be discussed with the Supervisor). Additional Information: There is a considerable amount of background information in the form of journal and conference publications, which will be available to the students working in this project. Also, in the ITEE School there are first generations of the working prototypes of microwave breast radar, which can be used to accomplish some of the tasks of the proposed project. The relevant information is covered in the past BE theses accomplished in 2007 and 2008. Collaboration with the ITEE PhD students working in this field is envisaged. In order to undertake this project the students should have background in EM fields and waves or microwaves, with the current GPA of 5.8 or above. | ||
Mikael Boden
Office: IMB
Phone: 52035
Email: mikael@itee.uq.edu.au
1 - Algorithms for discovering molecular signatures
| Supervisor: | Mikael Boden | Project ID: | 1 |
|---|---|---|---|
| Research Group: | Complex and Intelligent Systems Group | Max. students: | 1 |
| Discipline(s): | Num. students signed up: |
1 | |
| Prerequisite(s): | Programming skills, data structures and algorithms, and an interest in artificial intelligence/machine learning | ||
| Description: | My research group uses machine learning algorithms to make biological discoveries from large data sets resulting from genomic sequencing and wet-lab experiments. This project will look at the use of "molecular signatures" from proteins, to identify interactions between them. Molecular signatures can incorporate essential structural properties and make them accessible to state-of-the-art machine learning algorithms (e.g. support-vector machines). You will focus on the implementation of a "signature" function and use our implementation of a support-vector machine (java), to evaluate the approach on some scientifically interesting protein interaction data. | ||
3 - Pairing interacting proteins using sequence alignment
| Supervisor: | Mikael Boden | Project ID: | 3 |
|---|---|---|---|
| Research Group: | Complex and Intelligent Systems Group | Max. students: | 1 |
| Discipline(s): | Num. students signed up: |
2 | |
| Prerequisite(s): | Interest in data structures and algorithms, artificial intelligence and machine learning | ||
| Description: | This project will develop and explore a basic sequence alignment method for identifying the protein sub-sequences (of amino acids) that physically interact. First, we collect stats for amino acids that pair up in experimentally confirmed structures. The statistic is then used to create amino acid scoring matrices. The scoring matrices are used to find alignments between pairs of sequences, indirectly illustrating where on the sequences interactions are likely to be found. Finally, the method is evaluated by using it to predict interfaces on proteins for which experimental validation is available. The outcome thus includes a novel method that assists biological research to uncover underlying sequence features of large-scale molecular interaction networks. | ||