Machine learning algorithms for analysis of Infrared spectroscopic imaging of tissue samples for toxicology studies in rats.
Currently, tissue biopsy samples are examined by a trained pathologist using conventional optical brightfield microscopy. The disadvantages of this include the laborious and time consuming manual examination of the tissue under the microscope, delays between biopsy harvesting and diagnosis, inter- and intra-observer errors, the expense of using trained pathologists and the fact that many lesions turn out to be benign.Infrared (IR) chemical imaging, or spectral histopathology, has shown great promise in recent years as a method to complement current histopathology practice. In contrast to conventional microscopy, a digital image of the tissue on a slide, exposed to infrared light, is generated and is analysed by a computer. IR spectroscopy utilises radiation to excite molecular vibrations to produce unique spectral images based on the protein, carbohydrate and lipid content of the tissue. Using (IR) spectroscopy, in combination with machine learning and pattern recognition techniques, differences between diseased and non-diseased tissues can be determined in an automated and objective process.
Although most work has been focused on clinical diagnosis, the application of machine learning techniques, including deep learning, could have a significant impact in reducing the number of animals used in animal studies. In this PhD project, based in the group of Prof. Peter Gardner (http://gardner-lab.com/), the student will develop and apply novel machine learning algorithms to investigate biopsy tissue examined using both brightfield microscopy and IR hyperspectral imaging.
The PhD student should ideally have a strong background in computer science and data analysis/data engineering or other related discipline. The project is focused on the machine learning analysis aspects of the spectroscopic hyperspectral imaging data. The project is interdisciplinary so the student should be able to work independently but also interact with chemists, biologists and pathologists and our industrial partners who will be involved in this research. If you wish to discuss this project further please contact [Email Address Removed]
The PhD student should ideally have at least a 2.1 honours degree in computer science and data analysis/data engineering, related discipline.
Applications are open to UK and EU students. Self-funded overseas students may also apply
Industrial funding is available for this project. A £3,000 top-up to the normal stipend is available for an exceptional candidate.
How good is research at University of Manchester in Aeronautical, Mechanical, Chemical and Manufacturing Engineering?
FTE Category A staff submitted: 33.90
Research output data provided by the Research Excellence Framework (REF)
Click here to see the results for all UK universities