Cookies on this website

We use cookies to ensure that we give you the best experience on our website. If you click 'Accept all cookies' we'll assume that you are happy to receive all cookies and you won't see this message again. If you click 'Reject all non-essential cookies' only necessary cookies providing core functionality such as security, network management, and accessibility will be enabled. Click 'Find out more' for information on how to change your cookie settings.

 

Funded project – DPhil in Translational Health Sciences

 

Supervisors: Professor Catherine Pope; Professor John Powell; Dr Abi McNiven

Potential applicants should contact Dr McNiven to discuss the project in the first instance.

 

Application deadline - 12:00 midday (UK time) Friday 29th July 2022

 

The Department

The Nuffield Department of Primary Care Health Sciences is made up of a diverse mix of people and we actively promote a culture that supports a healthy work–life balance. The University is committed to increasing the number of postgraduate students from under-represented groups at Oxford and in recent years has introduced a number of pioneering initiatives to improve access to graduate study. The department is an enthusiastic participant in a continuing pilot to improve the selection procedure for graduate applications, in order to ensure all candidates are evaluated fairly.

 

Project description

This project will explore perspectives and concerns about racial bias in the development, uptake and implementation of digital health technologies/artificial intelligence for skin disease. As a related but secondary objective in addressing ethnic health disparities, it will also examine the professional issues in the adoption of these disruptive technologies.

Digital tools are being adopted in the context of wider policy initiatives to harness digital health technologies to deliver safe, effective and patient-centred care, at low marginal cost. Skin lesions have been seen to lend themselves to automated approaches whereby images can be assessed using algorithms and artificial intelligence (AI), and there are a growing number of dermatology apps aimed directly at consumers (patients) and/or professionals.

In this disrupted landscape of digital dermatology, several challenges may hinder the successful adoption of novel technologies and may in fact lead to patient harm and other unintended consequences. These include: (1) the potential ‘algorithmic’ racial bias in how digital dermatology tools perform; and (2) the threats to the profession of dermatology engendered by the automation of tasks previously conducted by a dermatology specialist (specifically the visual and tactile assessment of lesions). Both these issues are of concern to patient groups, and to professional societies such as the British Association of Dermatologists.

It is only in recent years that mainstream attention has been called to the predominance of white skin in photographs and symptom descriptions in dermatology textbooks and medical curriculum. Automated tools and AI approaches to skin lesions require the software to ‘learn’ from training datasets and bias occurs when, as in the case of medical textbook photographs, the training data is unrepresentative of all skin types. New visualisation techniques may be required to address lower contrast or visibility of the underlying vasculature on non-white skin. Understanding public, patient and clinician perspectives about the potential for racial bias in such technologies is key to addressing inequities in digital healthcare.

There is also a need to better understand the impacts of digital technologies on dermatological practice, e.g. the ‘uberisation’ of dermatology whereby consumers (patients) may self-diagnose with no contact with health services. We know that many dermatologists already have significant concerns that digital image based tools are reductive and compromise effectiveness and safety of care. This project will examine how digital tools support and/or challenge the role of clinicians providing dermatology care, and develop strategies that ensure that digital tools are used in partnership with traditional approaches, for example to support timely diagnosis of people who need urgent care. 

We anticipate that the proposed DPhil will draw on sociological perspectives and that the project design and methods will be qualitative or mixed methods (e.g. using interviews, focus groups, ethnography and evidence synthesis). Patient and Public Involvement and Engagement will be embedded throughout, from design to dissemination. Experts in digital health and in dermatology will also be engaged. Communication will be key to the outputs of the DPhil, including to raise awareness of findings amongst clinicians and other users of AI of the need to take account of diversity, including through using diverse training sets.

 

Funding

This is project is a funded studentship in association with Oxford PharmaGenesis and Green Templeton College.

The successful candidate will have their tuition fees covered (Home or Overseas level) and receive an annual stipend of at least £16,077.