Developing tomographic 3-D ultrasound for cardiovascular screening

Research question: Can, in combination with patient risk factors, tomographic 3-D ultrasound (tUS) accurately/automatically detect features of plaque instability in asymptomatic carotid disease (CAD), therefore identify patients at increased risk of stroke/cardiovascular events (CVE), be good value-for-money for the NHS by measuring atherosclerotic burden/volume.

Background: Cardiovascular disease (CVD) is the leading cause of death in the Western World; costing the UK economy £29 billion/year.

As the long-term effects of CVD are largely preventable, early detection by population screening could massively reduce the frequency and cost of CVE.

We know that screening the general population is not cost effective at a stroke rate of below 1.7% but in high-risk groups with greater disease prevalence it may be. CT and MR angiography will never be appropriate for screening because they involve radiation or are not quick.

Alternatively, standard duplex ultrasound can quickly detect asymptomatic carotid disease but requires highly trained Vascular Scientists, of which there is a national shortage.

Tomographic 3-D ultrasound (tUS) is quicker, requires less skill, and with development, could be deployed in the community. Principally, screening practitioners could be trained within a matter of weeks. However, measuring arterial disease is slow and currently still requires Vascular Scientists.

Aims: To develop and understand the potential clinical and cost effectiveness of automating tUS to detect asymptomatic carotid artery disease (CAD) plus measure atherosclerotic burden in the clinical setting to predict patients who are at risk of stroke.

Methods: I will measure carotid atherosclerotic burden (Carotid Plaque Volume or CPV) in 2570 high-risk patients from peripheral vascular or abdominal aortic aneurysm clinics. The prevalence of CAD is as high as 25%. Each patient will have tUS imaging of their aorta and carotid arteries to train AI to calculate CPV.

CPV measured using artificial intelligence (AI) machine learning techniques will then be validated against manual measures on MRI and tUS imaging. Models able to predict CVE based on CPV and other established risk factors will then be developed.

The cost-effectiveness/value-for money of utilising tUS for the NHS will be calculated under supervision from the Manchester Centre for Health Economics.

Timelines: The research will take place over 3-years within the world centre of excellence for tUS vascular imaging alongside world leading experts.

All patients will be recruited over the first 12 months and then followed up for 2-years before history on CVE is recorded. In project year 2 the measurements of atherosclerotic disease will be performed and automated. A risk prediction model and value-for-money will be calculated in year 3.

Impact and dissemination: My project will deliver an automated CVD screening tool ready to be validated within a national screening programme that is proven to be cost-effective for the NHS.

It will require little training to be utilised by screening practitioners and identify individual patients at high-risk of stroke.

Results will be presented at international vascular conferences and published open access as well as advertised via social media, all with PPI involvement.