_PMaguire2: Pattern recognition techniques for microsensors to diagnose disease from human breath
We have developed a new sensor method based on microplasmas which can operate at normal pressure. These opto-electrical based devices have been shown to detect trace gases down to the level of parts per trillion. For this reason they have possible application in the detection of specific disease markers. It has been known for centuries, that breath can provide strong clues to the state of a person's health. Recently there has been exciting discoveries indicating the potential to detect asthma, diabetes and possibly even lung cancer. However, to actually detect disease and act an early warning system, requires sensors and associated pattern recognition techniques that can accurately find very small disease signals within the much higher intensity normal patterns. We have developed a new optical sensor capable of capturing the very large arrays of data necessary to generate accurate recognition of disease states by developing suitable pattern recognition algorithms. These algorithms may be based on techniques such as neural nets, cluster analysis, principal component analysis, genetic algorithms, among others.
If you are interested in any of my projects, please email me for more details.
To apply for any of these PhD projects, FOLLOW GUIDELINES FOR ONLINE APPLICATION HERE www.compeng.ulster.ac.uk/rgs/guideForApplicants.php
Copy and Paste title and Description of this project into the application form (section: Referees & Research Proposal). You may also any two of my other projects in this application form - just include titles
The project details below are as provided as the initial outline. You will have the opportunity to discuss the exact project details with me and together we will develop a research programme to suit your expertise and preferences.
Qualifications; First degree in any of the following: Computer Science, Electronic Engineering, Physics, Physical Sciences, Materials Science, or similar subject area.
Subject For students interested in any of the following areas: Nanotechnology, Sensors, Pattern Recognition, Computation, Microplasmas, Biomedical Diagnostics, Microfabrication
Personnel Involved
First Supervisor: Maguire, PD Prof
Second Supervisor: Mariotti, D Dr
Collaboration: This project does not involve collaboration with another establishment
Synopsis:
Recently there has been exciting discoveries indicating the potential to detect asthma, diabetes and possibly even lung cancer from breath measurements. However, to actually detect disease and act an early warning system, requires sensors and associated pattern recognition techniques that can accurately find very small disease signals within the much higher intensity normal patterns. We have developed a new optical sensor capable of capturing the very large arrays of data. With this data we now need to develop software algorithms that will recognise the signature patterns of disease. These pattern recognition algorithms may be based on techniques such as neural nets, cluster analysis, principal component analysis, genetic algorithms, among others.