AI4Health Vision

Our vision is to establish a world leading centre for PhD training of next-generation innovators in Artificial Intelligence (AI) applied to Healthcare. We take the view that AI is ultimately a question about how to realise artificial systems that solve problems presently requiring human intelligence to solve (e.g. problems solved by clinicians, nurses and therapists). The contemporary answer to this question machine learning, i.e. learning how to solve the problem from experience instead of programming it. Learning in turn requires data, which extends our view of AI to encompass also data science and data engineering.

AI in our centre implies an entire system that solves a practical (healthcare) problem, this may require combining different algorithmic and representational approaches that encompass not only machine learning but also causal inference & symbolic reasoning, natural language processing, computer vision etc. Tackling healthcare challenges requires learning how to bridge our understanding of the clinical language and methodologies, the regulatory, legal and ethical frameworks of healthcare with core AI technical skills. Our training outcomes are PhDs that have learned to move fluidly across the disciplinary AI/healthcare boundaries and can develop and implement deployable solutions - therefore we have paired with two BRCs, three NHS Trusts and the NHS Digital Academy, among other partners to deliver unique healthcare training.

Our bespoke training program will enable our CDT students to emerge from their PhD as future leaders in AI industry and healthcare able to bring about transformative change by equipping them with the technical AI skills, an understanding of regulation and embedding them in an ecosystem of industry, accelerator, regulator and healthcare partners.

Our PhD training has 4 broad healthcare themes that leverage AI in healthcare: (1) Making healthcare provision more efficient and effective by increasing the productivity of doctors and nurses or patients through assistive technology; (2) Developing AI based diagnostics & monitoring that can e.g. detect disease earlier and monitor health with higher precision; (3) AI-based decision support systems that will e.g. free up doctors’ and patient’s time to focus on key challenges or optimising the delivery or development of treatments (4) AI that accelerates patient-centric drug and treatment development. These CDT’s research themes arise from our engagement with our healthcare partners (Imperial's three affiliated NHS trusts, Imperial Healthcare NHS Trust, Royal Marsden NHS Foundation Trust), technology companies and patient organisations. We thus look forward to seeing the full breadth of our partners medical capabilities reflected in our potential topics.

Supervisors

Each PhD student will have (at least) two supervisors - one AI supervisor and one Healthcare supervisor that will be in two different Faculties or Departments. Our view is to prefer supervisor pairs that are from two different faculties but we also allow supervisors from the same faculty in two different Departments (this rule also accommodates the increasing number of colleagues with joint appointments across two faculties).

The supervisor pool currently comprises:

Additional healthcare supervisors can be from clinicians in:

with or without Imperial Honorary Contracts (we can sponsor an Imperial College Honorary Contract as needed through the kind support of the Academic Health Sciences Centre)

We run an open CDT model, so other Departments at Imperial or individual supervisors can join, subject to them subscribing to our CDT’s financial model. Please contact our CDT Management team by emailing ai4health-management@imperial.ac.uk .

Admissions & Project Model

PhD students are invited to apply via the Imperial College application portal (also see our CDT information on How to Apply). Applications are reviewed and ranked by the CDT’s Admissions Panel (composed of academics, such as directors of postgraduate research or deputies of participating departments). The CDT will then rank the best X students and offer them to choose from a set of Y projects, where Y>>X and X up to the confirmed number of places for each academic year .

The CDT we will not rank projects per se but instead the CDT Research Board will check for AI remit and feasibility. Shortlisted applicants will get access to a pool of available projects and rank their preferred projects (up to 3).  Candidates will also have opportunities to discuss the projects with the supervisors and will also be interviewed by the potential supervisors. Alternatively, the CDT panel might call shortlisted candidates for interviews. The CDT aims to return decisions within 2-3 months of applying.

The AI4Health admissions team is reachable under ai4health-admissions@imperial.ac.uk .

PhD Milestones

The formal PhD milestones (e.g. ESA, LSR) are structured as cohort-based activities, and thus the CDT will run and organise these.

The CDT has a student progression/monitoring process in line with College regulations and thus are very similar to the 11 participating Departments (e.g. ESA is due around month 11, exit before month 12 deadline; LSR by month 24). Similarly, as required by UKRI we have an exit degree (an MRes) which students who want to or have leave in the first year can obtain .