Job description

Medicine is undergoing a data revolution, with AI being the engine of change. To achieve this, algorithms commonly require a significant amount of labelled data, a process which is very time consuming and expert-user intensive. To facilitate and speed-up this labelling process, human raters can make use of AI algorithms to assist the annotation process. This can be done by suggesting an initial segmentation to be curated or improved by a human rater, by selecting particular slices or subjects that are hard to segment, all with the aim of maximising the AI algorithm labelling accuracy while minimising user interaction time.

The environment:

The London Medical Imaging & AI Centre for Value-Based Healthcare is a consortium of academic, NHS and industry partners led by King s and based at St Thomas Hospital. Our diverse research teams are training sophisticated artificial intelligence algorithms from a vast wealth of NHS medical images and patient pathway data to create new healthcare tools. For patients, these will provide faster diagnosis, personalised therapies and effective screening across a range of conditions and procedures.

Through a focus on our experience in value-based healthcare we are examining how AI can be used to optimise triage and target resources to deliver significant financial savings for the NHS and healthcare systems overall. The centre has been established as part of the UK Government s Industrial Strategy Challenge Fund, delivered through UK Research and Innovation.

The purpose of this role:

This is an exciting opportunity for an enthusiastic deep learning researcher to push the boundaries of human-AI interaction, active learning, image segmentation, object detection, and image classification.

This role will be part of the AI4VBH Centre and will help deliver on a data labelling infrastructure, comprised of visualisation/contouring software and AI models. More specifically, this post will develop new algorithms and associated software stack to enable AI-assisted annotation for the problems of image segmentation, object detection and image classification. The post holder will focus on technical algorithmic developments such has using model uncertainty for active-learning based image/slice prioritisation, AI-based contouring (similarly to grab-cut), and general-purpose model pretraining to bootstrap segmentation, object detection and classification tasks on many different body parts and image modalities.

These algorithms shall be integrated and deployed into the AI4VBHC centre infrastructure as a proof-of-concept, utilising the data management and computational infrastructure of the AI centre.

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

This is a full-time post

Key responsibilities

The successful applicant will be responsible for developing active-learning based segmentation and object-detection algorithms that interact with a human rater.

Models and algorithms shall build on the MONAI (monai.io) and Pytorch machine learning stack, should make use of XNAT as an imaging data retrieval database, and should be made available to the community as open source, continuing the AI centre s open source tradition.

The applicant shall also engage and further develop research relationships with key AI-centre-related hospital partners (e.g. KCH, GSTT, and UCLH), and industrial partners (e.g. NVIDIA, Siemens, and several SMEs).

The applicant should ideally have some knowledge and experience of:

1. Medical image analysis

2. Deep Learning

3. Data science

4. Statistical modelling

The position would appeal to a candidate with strong software development skills, including:

1. Classic machine learning and deep learning packages (e.g. pytorch, )

2. Data science and statistical modelling packages (Pandas, SciPy/Statsmodels)

3. Container orchestration tools (Kubernetes, Docker)

4. Web technologies (e.g. RESTfull APIs, python-based web servers)

Previous experience with large-scale computing services would be useful but not essential. Strong communication skills are required to work with researchers from other disciplines, such as clinical end-users, and industry collaborators.

The candidate is also expected to:

- work in close collaboration with KCH, GSTT, and UCLH colleagues

- work in close collaboration with our key industrial partners, NVIDIA, Siemens and other SMEs

- work towards a common software stack

Experience working with interdisciplinary teams of engineers and clinicians will be valued. A strongly independent applicant is required who will need to work well with inter-disciplinary teams.

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

GRADE 6

Essential criteria

1. PhD awarded in Mathematics, Engineering or Computer Science or PhD in Mathematics, Engineering or Computer Science near completion*.

2. Undergraduate or higher degree in engineering, applied maths or computer science

3. Higher language computer programming (e.g. Python)

4. General machine learning experience

5. Scientific / Medical Writing

6. Machine learning libraries (SciKit Learn, pytorch, )

7. Interest in medical imaging

8. Ability to work and develop clinical and industrial relationships

9. Ability to work calmly under pressure and act on initiative

Desirable criteria

1. Image segmentation and object detection

2. HPC computing services (SGE-like batch-queuing system, Kubernetes, Docker, Containers)

3. Independent and interdisciplinary researcher

GRADE 7 - As above PLUS:

Essential criteria

1. Substantial post-doctoral research experience in Medical Image analysis and Deep Learning

2. Demonstrable experience with industry collaboration