Job description

Dr Emma C. Robinson of King’s College London is looking for a talented research software engineer for a new grant looking to develop novel radiological support tools which significantly improve the precision with which brain abnormalities may be detected in individual brains. 

The project will involve extension of MSM (Multimodal Surface Matching) – a tool for brain image registration that has already greatly improved understanding of brain organisation and development (e.g. Glasser MF Nature 2016, Garcia K PNAS 2018) - in a way that will more explicitly account for individual variability in brain shape and organisation. This will support much more accurate comparisons of brain shape, function and microstructure across large groups of individuals.

In this way, leveraging big open data sets such as the UK Biobank (> 40,000 datasets of healthy ageing adults), the Human Connectome Project (HCP > 1100 young healthy adults), and the developing HCP (dHCP < 1000 neonates and fetsuses); the objective is to significantly improve diagnosis and localisation of neurological pathologies by building precise normative models of healthy brain variation, from which abnormalities may be detected as outliers. This approach will be validated on focal epilepsy localisation, and detection of biomarkers of mental health and neurodevelopmental impairment.

This project will be co supervised by translational experts Dr Jonathan O’Muircheartaigh and Prof. Serena Counsell; and through collaboration with the Wellcome Centre for Integrative Neuroimaging (WIN) Oxford University, and the Washington University St Louis, thereby supporting ongoing development of the HCP, dHCP and UK Biobank datasets.

This post will be offered on an a fixed-term contract for 2 years

This is a full-time post

Key responsibilities

• Improving speed and parallelisation of current MSM code base by either 1) optimising current C++ code base; 2) Restructuring in Python and Pytorch
• Extending the codebase to design and implement a novel groupwise registration approach.
• Developing new machine learning tools for clustering brains by common patterns of cortical shape and organisation
• Developing novel normative models of population variation through which abnormalities attributable to focal epilepsy, neurodevelopmental impairment and mental health disorders may be found. Generative deep learning techniques will also be considered.
• Image processing for open science frameworks – the Human Connectome Project (HCP), developing Human Connectome Project (dHCP) and UK Biobank.
• Publishing in journals and major biomedical conferences.
• Code maintenance, documentation and user support.

The above list of responsibilities may not be exhaustive, and the post holder will be required to undertake such tasks and responsibilities as may reasonably be expected within the scope and grading of the post.

Skills, knowledge, and experience

Research Software Engineer

Essential criteria

1. PhD or Graduate degree (MSc or PhD) in a computing/engineering subject OR equivalent industrial experience in software development.
2. Experience in software engineering of large projects
3. Strong coding ability in Python and C++
4. Proven ability to convert research papers or maths into code
5. Experience of managing code releases and provide user support
6. Knowledge of version control systems e.g. GitHub
7. Knowledge of at least one of numerical optimisation, machine learning or graphics/accelerated programming
8. Ability to work collaboratively with people from a variety of technical and non-technical backgrounds
9. Track record in working proactively and independently (at home and in the office)

Desirable criteria

1. Experience in image analysis, particularly biomedical
2. Experience with machine learning specifically computer vision and/or normative modelling
3. Expertise in discrete/combinatorial optimisation

Research Associate

Essential criteria

1. PhD in Biomedical Engineering, Computing or a related numerate discipline*
2. Experience in software engineering of large research projects in at least one of the following: Python, PyTorch and C++
3. Experience in developing novel methods for analysing imaging data, preferably biomedical
4. Proven ability to convert research papers and maths into code
5. Knowledge of at least one of numerical optimisation, machine learning or graphics/accelerated programming
6. Experience in publishing in major journals and/or biomedical conferences
7. Ability to work collaboratively with people from a variety of technical and non-technical backgrounds
8. Evidence of public release of code.
9. Track record in working proactively and independently (at home and in the office)

Desirable criteria

1. Experience with image registration

2. Knowledge of normative/generative modelling

3. Expertise in discrete/combinatorial optimization

4. Knowledge of version control systems e.g. GitHub