Computational Image processing: Super-resolution restoration of microscopy images and its application in biological and biomedical imaging

Super-resolution microscopy is an emerging field that tries to overcome the resolution limit of conventional microscopy. This project proposes to take a multi-disciplinary approach to this problem by combining the fields of image processing with microscopy imaging to develop a novel super-resolution restoration method. Our research group has recently been developing such a super-resolution imaging technique, called translation microscopy (TRAM), in which a super-resolution image can be restored from multiple diffraction-limited (low resolution) images recorded from standard microscopes. This is basically an inverse problem, which retrieves high resolution images from diffraction-limited observations. It is solved by the model-based approach, referred to as image restoration. Our preliminary results have demonstrated that TRAM is capable of improving lateral resolution by 7-fold, delivering a multi-colour image with ~30nm resolution [see our recent publication: Zhen Qiu, et al, Translation Microscopy (TRAM) for super-resolution imaging, Scientific Report 6, Article number: 19993 (2016)]. This project aims to further develop this methodology, delivering a completely novel software application that would for the first time enable quantitative multi-colour super-resolution imaging. In this project we are particularly interested in its application to cancer imaging deep in living tissue. This work, in collaboration with biologic and biomedical colleagues, aims to improve our understanding on how cells behave in their natural environment, and quantifying changes in cell behaviour and the underlying mechanisms that mediate the changes. Furthered development of the TRAM methodology for cancer imaging has the potential to transform pre-clinical in vivo cancer research.

This project aims to further develop this methodology, delivering a completely novel software application that would for the first time enable quantitative multi-colour super-resolution imaging. In this project we are particularly interested in its application to cancer imaging deep in living tissue. This work, in collaboration with biologic and biomedical colleagues, aims to improve our understanding on how cells behave in their natural environment, and quantifying changes in cell behaviour and the underlying mechanisms that mediate the changes. Furthered development of the TRAM methodology for cancer imaging has the potential to transform pre-clinical in vivo cancer research.

The candidate should have a good honours degree (1st or upper 2nd class, or equivalent) in mathematics, physics, computer sciences or engineering and a strong desire to develop his/her research career in image processing and analysis particularly for biological and biomedical applications. Previous research experience in medical imaging or a postgraduate degree in image processing will be an advantage but not essential. A successful candidate must have a strong interest in multi-disciplinary researches.