Applications are invited for a PhD studentship in high-performance medical image analysis, under the supervision of Dr Bernhard Kainz in the Biomedical Image Analysis Group, Department of Computing at Imperial College London. The PhD position is fully funded for UK/EU candidates. Overseas applicants may apply but will be required to show evidence that the difference between the overseas fees and UK/EU fees will be covered by other funding. The start date is April or October 2016.
Motion correction is a core problem for the evaluation of medical images from various imaging modalities like Magnetic Resonance Imaging (MRI), Ultrasound, and Computed Tomography (CT). Involuntary movements of organs during the scan or the motion of fetuses inside the uterus cannot be avoided during image acquisition. Therefore, computational methods are currently used to retrospectively compensate for unavoidable motion artefacts. However, the state-of-the-art algorithms that are currently deployed are mainly based on image intensities or simple motion models and they only work for a limited amount of random motion or for highly predictable periodic motion patterns. The aim of this project is to explore machine learning methods to automatically integrate domain expert knowledge about the semantic meaning of different areas of the image. Expert knowledge can be encoded in average anatomical models from a large image cohort of the organ of interest. The development of these models for selected applications like cardiac MRI, fetal MRI and fetal ultrasound will be one of the core parts of this project. Ultimately, automatic semantic understanding of the contents of a dataset will allow to develop a generic approach for random and periodic motion correction. High-performance computing methods based on commodity hardware like graphics processing units (GPUs) will allow to apply these methods already during the image acquisition process.
The successful applicant will be expected to develop their expertise in the interdisciplinary area of the project as well as to contribute to research and to the related publication activities.
Prospective applicants will have a good first degree (or equivalent qualification) in Computer Science, Mathematics, Physics or another engineering-related discipline. Good programming skills (C-like languages and scripting) are required, knowledge of parallel computing, machine learning, mathematical modelling and optimization is desirable. We are looking for highly motivated applicants with excellent interpersonal, written and oral communication skills and enthusiasm for exposure to a diversity of scientific projects.
To apply for this position, please email a single PDF file including: a cover letter describing your interests and research experience, your CV, a copy of transcripts, and names and contact information of two references to Dr Bernhard Kainz ([email protected]
) with the subject line ‘PhD application 2016’. Although applications will be accepted until the position is filled, early applications will be given priority.