PhD Studentship: Reducing Scan Time and Radioactive Exposure in Nuclear Imaging Using Swarm Intelligence

This scholarship will support the PhD candidate to carry out frontier research in swarm led Tomographic Reconstruction (TR), which is the process of inferring the internal structure of an object from the projected images cast by penetrating radiation. TR plays an essential computational role in all medical imaging procedures (X-Ray CT, PET, MRI, Nuclear Medicine, and ultrasound). It also has wide application in industry, science, and mathematics. The reconstruction problem is similar to inferring the 3D structure of an object from its shadow or filling a grid of numbers given the row and column sums (a style of Japanese crossword). 

The PhD candidate will work under the supervision of Dr. Mohammad Majid al-Rifaie at the University of Greenwich, London. The supervisory team includes Dr. Hooman Oroojeni (University of Greenwich), along with Dr. Tim Blackwell (Goldsmiths University of London) and Dr. Bal Sanghera (Paul Strickland Scanner Centre, Mount Vernon Hospital). 

We are looking for an outstanding student to work with the team.

The research will pioneer the development of new class of population-based algorithms tailored for few-view tomographic reconstruction. These techniques include: dispersive flies optimisation (DFO), particle swarm optimisation (PSO) and differential evolution (DE). The aim of the project is to evolve a tomographic reconstruction process founded on swarm intelligence algorithms and capable of clinically acceptable reconstructions. The proposed method will be comparable to conventional techniques for normal images but will surpass conventional techniques for situations where the number of projections is reduced. This work is substantiated by a body of work published by the supervisors. 

The short-term objective is to demonstrate feasibility and lay the foundation of a swarm-plausible framework over experimental phantoms such as the clinically important Shepp-Logan phantom and others. 

The team's long-term goal is to create a framework capable of reconstructing images from real-world clinical projections under few-view scenarios. This work benefits from the existing literature on swarm intelligence, optimisation in high dimensional search spaces, and traditional and novel techniques in nuclear imaging. The ultimate goal is a reduction in scan time and radioactive exposure.

Funding Details

Bursary available (subject to satisfactory performance):

Year 1: £17,668 plus London weighting where applicable (FT) or pro-rata (PT) Year 2: In line with UKRI rate Year 3: In line with UKRI rate.

In addition, the successful candidate will receive a contribution to tuition fees equivalent to the university’s Home rate, currently £4,596 (FT) or pro-rata (PT), for the duration of their scholarship. International applicants will need to pay the remainder tuition fee for the duration of their scholarship.

This fee is subject to an annual increase.

Closing Date: midnight UTC on 19th June 2023

Advert information