Breast cancer is the most common type of cancer worldwide and is estimated to affect one in nine women over their lifetime.1,2 Survival outcomes associated with breast cancer are greatly improved with early diagnosis, and consequently the National Health Service (NHS) launched its Breast Screening Programme in 1988 to improve early detection rates. The programme currently screens women aged 50-70 years every three years, with approximately 1.3 women attending annually. The screening programme uses X-ray images (mammograms) to detect potentially cancerous lesions requiring further investigation. The mammogram images are currently reviewed independently by two trained human readers, with arbitration of discrepancies by a third reader.3
The process of manual reading is resource intensive, and as demand and workforce pressures, exacerbated by COVID, continue to affect breast cancer screening,4 there is currently great interest in the potential for automation, using artificial intelligence (AI) algorithms, to improve the efficiency of the reading process and service as a whole. There are several potential implementation models by which AI could achieve this: 1) replacing one of the manual readers; 2) acting as a support tool for manual readers, to optimise their workflow and throughput; 3) acting as a sole reader; or 4) acting as an arbitration grader. To inform decisions on whether/how to utilise AI in the screening programme, it is vital to consider the expected costs and consequences of the different implementation models compared to standard practice.5 There are well established decision analytic modelling techniques to support the economic evaluation of diagnostic and screening technologies. However, there are some specific challenges relating to evaluation and implementation of AI technology that need careful consideration. These include but are not limited to 1) variability in technology performance over time, as algorithms continue to learn and/or image capture technology changes and affect performance; 2) identifying the full economic costs of embedding, maintaining and quality assuring AI technology in the IT systems in which they are to operate; and 3) the potential impact of practitioner and public perceptions on the likely acceptance and uptake of different AI implementation models. This PhD project will explore ways of addressing specific challenges for the evaluation and implementation of AI for breast cancer screening within an economic evaluation framework.
This project will suit a numerically skilled student with interests in economic evaluation and decision modelling, but also the use of mixed methods to understand barriers to implementation and/or public perceptions and preferences. Candidates should hold a post graduate qualification in health economics, medical statistics, or a related quantitative discipline.
You will join a multidisciplinary collaborative team with an international reputation in the areas of health economics and health services research, health data science applied to the development and evaluation of artificial intelligence algorithms, and clinical expertise in breast cancer screening.
Candidates should contact the lead supervisor (Dr Graham Scotland) to discuss the project in advance of submitting an application, as supervisors will be expected to provide a letter of support for suitable applicants. Candidates will be informed after the application deadline if they have been shortlisted for interview.
International applicants are eligible to apply for this studentship but will have to find additional funding to cover the difference between overseas and home fees (approximately £17,000 per annum)
- Formal applications can be completed online: https://www.abdn.ac.uk/pgap/login.php
- You should apply for the Degree of Doctor of Philosophy in Applied Health Sciences to ensure your application is passed to the correct team
- Please clearly note the name of the supervisor and exact project title on the application form. If you do not mention the project title and the supervisor on your application it will not be considered for the studentship.
- Candidates should have (or expect to achieve) a minimum of a First Class Honours degree in a relevant subject. Applicants with a minimum of a 2:1 Honours degree may be considered provided they have a Distinction at Masters level.
- General application enquiries can be made to [Email Address Removed]