EPSRC PhD Studentship: Novel Optics and AI Aproaches to Image Centre Root

Area

Chemical & Environmental Engineering

UK Other

Open Until Filled

ENG
308

Novel optics and AI approaches to image the centre of a live root for the first time. This exciting opportunity is based within the thriving Optics and Photonics Research Group in Faculty of Engineering which conducts cutting edge research spanning exploration to translation, with curiosity driven projects all the way through to application in the clinic.

We are seeking a PhD student that is motivated and enthusiastic and keen to push the boundaries of what is currently possible when imaging with an optical microscope. Combining the latest in optical developments with the recent surge in AI, this project aims to image the centre of a live intact root for the first time. Something that is currently not possible.

This project will address a long-standing issue in plant biology: the inability to image the centre of live, intact, plant roots. The ability to observe dynamic cellular processes at the centre of a live root for the first time will unlock entirely new lines of biological inquiry, crucial for areas such as sustainable agriculture and food security. Such an imaging system would allow for studies of a plant’s resilience to drought, salinity, and water logging, as well as responses to fungal infections and nanoparticle uptake.

It is very common that new optical microscopy techniques are developed to image mammalian tissue, and that these approaches are very slow to translate across to plant biosciences where the impact could be huge and as a result exciting opportunities get missed.

When we use light to image deep into complex samples there is a common problem that occurs – the light gets distorted and scattered by the structures present in the sample and as a result a nice quality focus and hence a nice image cannot be produced at depth into the sample. At Nottingham we have been working on this problem for several years and have developed methods that shape the incoming light with the equal but opposite distortion to that imposed by the sample to produce a high-quality image deep into the sample of interest.

Recently we have been using AI and machine learning to predict the distortion present and significantly speed up this correction process.

This PhD project will take the latest in AI-informed wavefront correction techniques and tailor them to imaging deep into plant roots. It will use a range of state‑of‑the‑art optical microscopes based in the Optics and Photonics Research Group in the Faculty of Engineering, plus those housed in Plant Biosciences at the Sutton Bonnington campus. Data sets will be generated using simulated and experimental data and these will be used to train networks to predict the common distortions that occur when imaging into plant roots.

From here we can either correct for these distortions using the hardware in the microscope or in software using reconstruction algorithms. This is an exciting multidisciplinary PhD project that promises to make cutting‑edge advances in all research areas involved.

This project combines practical hands‑on optics experimentation with training neural networks to develop the next generation of optical microscopes. You will have the opportunity to gain skills in optical instrumentation and imaging, AI and machine learning, and in plant biology and sample handling. Your base will be in the Optics and Photonics Group in the Faculty of Engineering and from here you will work with a team of academics and researchers across Engineering, Computer Science and the Biosciences.

You will be supervised by Amanda Wright (Optics and Photonics Research Group, Faculty of Engineering), Mike Somekh (Optics and Photonics Research Group, Faculty of Engineering), Mike Pound (Computer Vision, Computer Science Department), and Darren Wells (Plant and Crop Biophysics, School of Biosciences).

An enthusiastic, self‑motivated, resourceful student, who likes working as part of a team and is keen to take on a new challenge. An understanding of optics and/or machine learning is desirable but not essential,…