Deep learning for better cellular imaging
Join a cross-disciplinary team of researchers that seek to improve the imaging capabilities of scientists studying infection and reproductive biology.
This is a joint project led by physics and biophysics expert Professor Kishan Dholakia; Dr Danny Wilson from the University of Adelaide's School of Biological Sciences; and collaborators in the Faculty of Health and Medical Sciences.
Optical imaging is very exciting but has drawbacks for imaging across scales, such as visualising wide field of view with high resolution at the same time.
The physics/computing parts of this project will employ recent innovations in deep learning to retain the wide field of view and to computationally recover a super-resolved image.
Whilst a high optical resolution can be accomplished with the use of structured light fields that can carry extended spatial frequency information, many widefield systems suffer from the need to Nyquist sample the field of view, which fundamentally limits the dynamic range of the spatial length scale.
We can overcome this limit to resolution with the use of deep learning to recover high-resolution images from sub-sampled information. We will achieve this by using a generative adversarial network (GAN) model pre-trained on widefield fluorescence images. The aim will be to apply this deep learning image enhancement to explore key biology of malaria parasites and larger mammalian cells.
This project will have a focus on application and development of deep learning algorithms to improve imaging, but can be tailored to the level of expertise in coding that you may have experienced (e.g. PYTHON) to include more hands on imaging and biology as your interest level requires.
At the end of this project, you will have made significant steps towards creating a stand-alone module for image processing to allow a range of end users studying infection and reproductive biology to simply and efficiently improve their imaging capabilities.